Mark V. Williams, MD

Studies of the transition of care between inpatient and outpatient settings have shown that there are significant patient safety and quality deficiencies in our current system. The transition from the hospital setting to the outpatient setting has been more extensively studied than the transition from the outpatient setting to the inpatient setting. One prospective cohort study of 400 patients found that 1 in 5 patients discharged from the hospital to home experienced an adverse event, which was defined as an injury resulting from medical management rather than the underlying disease, within 3 weeks of discharge.1 This study also concluded that 66% of these were drug‐related adverse events, many of which could have been avoided or mitigated. Another prospective cross‐sectional study of 2644 patient discharges found that approximately 40% of the patients had pending test results at the time of discharge and that 10% of these required some action, yet the outpatient physicians and patients were unaware of these results.2 Medication discrepancies have also been shown to be prevalent, with 1 prospective observational study of 375 patients showing that 14% of elderly patients had 1 or more medication discrepancies and 14% of those patients with medication discrepancies were rehospitalized within 30 days versus 6% of the patients who did not experience a medication discrepancy.3 A recent review of the literature cited improving transitional care as a key area of opportunity for improving postdischarge care4

Lack of communication has clearly been shown to adversely affect postdischarge care transitions.5 A recent summary of the literature by a Society of Hospital Medicine (SHM)/Society of General Internal Medicine (SGIM) task force found that direct communication between hospital physicians and primary care physicians occurs infrequently (in 3%‐20% of cases studied), and the availability of a discharge summary at the first postdischarge visit is low (12%‐34%) and does not improve greatly even after 4 weeks (51%‐77%); this affects the quality of care in approximately 25% of follow‐up visits.5 This systematic review of the literature also found that discharge summaries often lack important information such as diagnostic test results, the treatment or hospital course, discharge medications, test results pending at discharge, patient or family counseling, and follow‐up plans.

However, the lack of studies of the communication between ambulatory physicians and hospital physicians prior to admission or during emergency department (ED) visits does not imply that this communication is not equally important and essential to high‐quality care. According to the Centers for Disease Control, the greatest source of hospital admissions in many institutions is the ED. Over 115,000,000 visits were made to the nation's approximately 4828 EDs in 2005, and about 85.2% of ED visits end in discharge.6 The ED is also the point of re‐entry into the system for individuals who may have had an adverse outcome linked to a prior hospitalization.6 Communication between hospital physicians and primary care physicians must be established to create a loop of continuous care and diminish morbidity and mortality at this critical transition point.

While transitions can be a risky period for patient safety, observational studies suggest there are benefits to transitions. A new physician may notice something overlooked by the current caregivers.712 Another factor contributing to the challenges of care transitions is the lack of a single clinician or clinical entity taking responsibility for coordination across the continuum of the patient's overall healthcare, regardless of setting.13 Studies indicate that a relationship with a medical home is associated with better health on both the individual and population levels, with lower overall costs of care and with reductions in disparities in health between socially disadvantaged subpopulations and more socially advantaged populations.14 Several medical societies have addressed this issue, including the American College of Physicians (ACP), SGIM, American Academy of Family Physicians, and American Academy of Pediatrics, and they have proposed the concept of the medical home or patient‐centered medical home, which calls for clinicians to assume this responsibility for coordinating their patients' care across settings and for the healthcare system to value and reimburse clinicians for this patient‐centered and comprehensive method of practice.1517

Finally, patients and their families or caregivers have an important role to play in transitions of care. Several observational and cross‐sectional studies have shown that patients and their caregivers and families express significant feelings of anxiety during care transitions. This anxiety can be caused by a lack of understanding and preparation for their self‐care role in the next care setting, confusion due to conflicting advice from different practitioners, and a sense of abandonment attributable to the inability to contact an appropriate healthcare practitioner for guidance, and they report an overall disregard for their preferences and input into the design of the care plan.1820 Clearly, there is room for improvement in all these areas of the inpatient and outpatient care transition, and the Transitions of Care Consensus Conference (TOCCC) attempted to address these areas by developing standards for the transition of care that also harmonize with the work of the Stepping up to the Plate (SUTTP) Alliance of the American Board of Internal Medicine (ABIM) Foundation.21 In addition, other important stakeholders are addressing this topic and actively working to improve communication and continuity in care, including the Centers for Medicare and Medicaid Services (CMS) and the National Quality Forum (NQF). CMS recently developed the Continuity Assessment Record & Evaluation (CARE) tool, a data collection instrument designed to be a standardized, interoperable, common assessment tool to capture key patient characteristics that will provide information related to resource utilization, clinical outcomes, and postdischarge disposition. NQF held a national forum on care coordination in the spring of 2008.

In summary, it is clear that there are qualitative and quantitative deficiencies in transitions of care between the inpatient and outpatient setting that are affecting patient safety and experience with care. The transition from the inpatient setting to the outpatient setting has been more extensively studied, and this body of literature has underscored for the TOCCC several important areas in need of guidance and improvement. Because of this, the scope of application of this document should initially emphasize inpatient‐to‐outpatient transitions as a first step in learning how to improve these processes. However, the transition from the outpatient setting to the inpatient setting also is a clear priority. Because the needs for transfer of information, authority, and responsibility may be different in these situations, a second phase of additional work to develop principles to guide these transitions should be undertaken as quickly as possible. Experience gained in applying these principles to inpatient‐to‐outpatient transitions might usefully inform such work.

Communication among providers and with the patients and their families arose as a clear priority. Medication discrepancies, pending tests, and unknown diagnostic or treatment plans have an immediate impact on patients' health and outcomes. The TOCCC discussed what elements should be among the standard pieces of information exchanged among providers during these transition points. The dire need for coordination of care or a coordinating clinician/medical home became a clear theme in the deliberations of the TOCCC. Most importantly, the role of the patients and their families/caregivers in their continuing care is apparent, and the TOCCC felt this must be an integral part of any principles or standards for transitions of care.

Methods

In the fall/winter of 2006, the executive committees of ACP, SGIM, and SHM agreed to jointly develop a policy statement on transitions of care. Transitions of care specifically between the inpatient and outpatient settings were selected as an ideal topic for collaboration for the 3 societies as they represent the continuum of care for internal medicine within these settings. To accomplish this, the 3 organizations decided to convene a consensus conference to develop consensus guidelines and standards concerning transitions between inpatient and outpatient settings through a multi‐stakeholder process. A steering committee was convened with representatives from ACP, SGIM, SHM, the Agency for Healthcare Research and Quality (AHRQ), ABIM, and the American Geriatric Society (AGS). The steering committee developed the agenda and invitee list for the consensus conference. After the conference was held, the steering committee was expanded to include representation from the American College of Emergency Physicians (ACEP) and the Society for Academic Emergency Medicine (SAEM).

During the planning stages of the TOCCC, the steering committee became aware of the SUTTP Alliance of the ABIM Foundation. The SUTTP Alliance has representation from medical specialties such as internal medicine and its subspecialties, family medicine, and surgery. The alliance was formed in 2006 and has been working on care coordination across multiple settings and specialties. The SUTTP Alliance had developed a set of principles and standards for care transitions and agreed to provide the draft document to the TOCCC for review, input, and further development and refinement.

Recommendations on Principles and Standards for Managing Transitions in Care Between the Inpatient and Outpatient Settings from ACP, SGIM, SHM, AGS, ACEP, and SAEM

The SUTTP Alliance presented a draft document entitled Principles and Standards for Managing Transitions in Care. In this document, the SUTTP Alliance proposes 5 principles and 8 standards for effective care transitions. A key element of the conference was a presentation by NQF on how to move from principles to standards and eventually to measures. This presentation provided the TOCCC with the theoretical underpinnings for the discussion of these principles and standards and how the TOCCC would provide input on them. The presentation provided an outline for the flow from principles to measures. First, there needs to be a framework that provides guiding principles for what we would like to measure and eventually report. From those principles, a set of preferred practices or standards are developed; the standards are more granular and allow for more specificity in describing the desired practice or outcome and its elements. Standards then provide a roadmap for identification and development of performance measures. With this framework in mind, the TOCCC then discussed in detail the SUTTP principles and standards.

The 5 principles for effective care transitions developed by the SUTTP Alliance are as follows:

• Accountability.

• Communication: clear and direct communication of treatment plans and follow‐up expectations.

• Timely feedback and feed‐forward of information.

• Involvement of the patient and family member, unless inappropriate, in all steps.

• Respect of the hub of coordination of care.

The TOCCC re‐affirmed these principles and added 4 additional principles to this list. Three of the new principles were statements within the 8 standards developed by the SUTTP, but when taking into consideration the framework for the development of principles into standards, the TOCCC felt that the statements were better represented as principles. They are as follows:

• All patients and their families/caregivers should have and should be able to identify their medical home or coordinating clinician (ie, practice or practitioner). (This was originally part of the coordinating clinicians standard, and the TOCCC voted to elevate this to a principle).

• At every point along the transition, the patients and/or their families/caregivers need to know who is responsible for care at that point and who to contact and how.

• National standards should be established for transitions in care and should be adopted and implemented at the national and community level through public health institutions, national accreditation bodies, medical societies, medical institutions, and so forth in order to improve patient outcomes and patient safety. (This was originally part of the SUTTP community standards standard, and the TOCCC moved to elevate this to a principle).

• For monitoring and improving transitions, standardized metrics related to these standards should be used in order to lead to continuous quality improvement and accountability. (This was originally part of the measurement standard, and the TOCCC voted to elevate this to a principle).

The SUTTP Alliance proposed the following 8 standards for care transitions:

• Coordinating clinicians.

• Care plans.

• Communication infrastructure.

• Standard communication formats.

• Transition responsibility.

• Timeliness.

• Community standards.

• Measurement.

The TOCCC affirmed these standards and through a consensus process added more specificity to most of them and elevated components of some of them to principles, as discussed previously. The TOCCC proposes that the following be merged with the SUTTP standards:

• Coordinating clinicians. Communication and information exchange between the medical home and the receiving provider should occur in an amount of time that will allow the receiving provider to effectively treat the patient. This communication and information exchange should ideally occur whenever patients are at a transition of care (eg, at discharge from the inpatient setting). The timeliness of this communication should be consistent with the patient's clinical presentation and, in the case of a patient being discharged, the urgency of the follow‐up required. Guidelines will need to be developed that address both the timeliness and means of communication between the discharging physician and the medical home. Communication and information exchange between the medical home and other physicians may be in the form of a call, voice mail, fax, or other secure, private, and accessible means including mutual access to an electronic health record.

  The ED represents a unique subset of transitions of care. The potential transition can generally be described as outpatient to outpatient or outpatient to inpatient, depending on whether or not the patient is admitted to the hospital. The outpatient‐to‐outpatient transition can also encompass a number of potential variations. Patients with a medical home may be referred to the ED by the medical home, or they may self‐refer. A significant number of patients do not have a physician and self‐refer to the ED. The disposition from the ED, either outpatient to outpatient or outpatient to inpatient, is similarly represented by a number of variables. Discharged patients may or may not have a medical home, may or may not need a specialist, and may or may not require urgent (<24 hours) follow‐up. Admitted patients may or may not have a medical home and may or may not require specialty care. This variety of variables precludes a single approach to ED transition of care coordination. The determination of which scenarios will be appropriate for the development of standards (coordinating clinicians and transition responsibility) will require further contributions from ACEP and SAEM and review by the steering committee.

• Care plans/transition record. The TOCCC also agreed that there is a minimal set of data elements that should always be part of the transition record. The TOCCC suggested that this minimal data set be part of an initial implementation of this standard. That list includes the following:

  The TOCCC discussed what components should be included in an ideal transition record and agreed on the following elements:

  The TOCCC also added a new standard under this heading: Patients and/or their families/caregivers must receive, understand, and be encouraged to participate in the development of the transition record, which should take into consideration patients' health literacy and insurance status and be culturally sensitive.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Clear identification of the medical home/transferring coordinating physician/emnstitution and the contact information.

• Patient's cognitive status.

• Test results/pending results.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Emergency plan and contact number and person.

• Treatment and diagnostic plan.

• Prognosis and goals of care.

• Test results/pending results.

• Clear identification of the medical home and/or transferring coordinating physician/emnstitution.

• Patient's cognitive status.

• Advance directives, power of attorney, and consent.

• Planned interventions, durable medical equipment, wound care, and so forth.

• Assessment of caregiver status.

• Communication infrastructure. All communications between providers and between providers and patients and families/caregivers need to be secure, private, Health Insurance Portability and Accountability Actcompliant, and accessible to patients and those practitioners who care for them. Communication needs to be 2‐way with an opportunity for clarification and feedback. Each sending provider needs to provide a contact name and the number of an individual who can respond to questions or concerns. The content of transferred information needs to include a core standardized data set. This information needs to be transferred as a living database; that is, it is created only once, and then each subsequent provider only needs to update, validate, or modify the information. Patient information should be available to the provider prior to the patient's arrival. Information transfer needs to adhere to national data standards. Patients should be provided with a medication list that is accessible (paper or electronic), clear, and dated.

• Standard communication formats. Communities need to develop standard data transfer forms (templates and transmission protocols). Access to a patient's medical history needs to be on a current and ongoing basis with the ability to modify information as a patient's condition changes. Patients, families, and caregivers should have access to their information (nothing about me without me). A section on the transfer record should be devoted to communicating a patient's preferences, priorities, goals, and values (eg, the patient does not want intubation).

• Transition responsibility. The sending provider/emnstitution/team at the clinical organization maintains responsibility for the care of the patient until the receiving clinician/location confirms that the transfer and assumption of responsibility is complete (within a reasonable timeframe for the receiving clinician to receive the information; ie, transfers that occur in the middle of the night can be communicated during standard working hours). The sending provider should be available for clarification with issues of care within a reasonable timeframe after the transfer has been completed, and this timeframe should be based on the conditions of the transfer settings. The patient should be able to identify the responsible provider. In the case of patients who do not have an ongoing ambulatory care provider or whose ambulatory care provider has not assumed responsibility, the hospital‐based clinicians will not be required to assume responsibility for the care of these patients once they are discharged.

• Timeliness. Timeliness of feedback and feed‐forward of information from a sending provider to a receiving provider should be contingent on 4 factors:

  This information should be available at the time of the patient encounter.

• Transition settings.

• Patient circumstances.

• Level of acuity.

• Clear transition responsibility.

• Community standards. Medical communities/emnstitutions must demonstrate accountability for transitions of care by adopting national standards, and processes should be established to promote effective transitions of care.

• Measurement. For monitoring and improving transitions, standardized metrics related to these standards should be used. These metrics/measures should be evidence‐based, address documented gaps, and have a demonstrated impact on improving care (complying with performance measure standards) whenever feasible. Results from measurements using standardized metrics must lead to continuous improvement of the transition process. The validity, reliability, cost, and impact, including unintended consequences, of these measures should be assessed and re‐evaluated.

All these standards should be applied with special attention to the various transition settings and should be appropriate to each transition setting. Measure developers will need to take this into account when developing measures based on these proposed standards.

The TOCCC also went through a consensus prioritization exercise to rank‐order the consensus standards. All meeting participants were asked to rank their top 3 priorities of the 7 standards, giving a numeric score of 1 for their highest priority, a score of 2 for their second highest priority, and a score of 3 for their third highest priority. Summary scores were calculated, and the standards were rank‐ordered from the lowest summary score to the highest. The TOCCC recognizes that full implementation of all of these standards may not be feasible and that these standards may be implemented on a stepped or incremental basis. This prioritization can assist in deciding which of these to implement. The results of the prioritization exercise are as follows:

• 1

  All transitions must include a transition record

• 2

  Transition responsibility

• 3

  Coordinating clinicians

• 4

  Patient and family involvement and ownership of the transition record

• 5

  Communication infrastructure

• 6

  Timeliness

• 7

  Community standards

Future Challenges

In addition to the work on the principles and standards, the TOCCC uncovered six further challenges which are described below.

Next Steps for the TOCCC

The TOCCC focuses only on the transitions between the inpatient and outpatient settings and does not address the equally important transitions between many other different care settings, such as the transition from a hospital to a nursing home or rehabilitation facility. The intent of the TOCCC is to provide this document to national measure developers such as the Physician Consortium for Performance Improvement and others in order to guide measure development and ultimately lead to improvements in quality and safety in care transitions.

Studies of the transition of care between inpatient and outpatient settings have shown that there are significant patient safety and quality deficiencies in our current system. The transition from the hospital setting to the outpatient setting has been more extensively studied than the transition from the outpatient setting to the inpatient setting. One prospective cohort study of 400 patients found that 1 in 5 patients discharged from the hospital to home experienced an adverse event, which was defined as an injury resulting from medical management rather than the underlying disease, within 3 weeks of discharge.1 This study also concluded that 66% of these were drug‐related adverse events, many of which could have been avoided or mitigated. Another prospective cross‐sectional study of 2644 patient discharges found that approximately 40% of the patients had pending test results at the time of discharge and that 10% of these required some action, yet the outpatient physicians and patients were unaware of these results.2 Medication discrepancies have also been shown to be prevalent, with 1 prospective observational study of 375 patients showing that 14% of elderly patients had 1 or more medication discrepancies and 14% of those patients with medication discrepancies were rehospitalized within 30 days versus 6% of the patients who did not experience a medication discrepancy.3 A recent review of the literature cited improving transitional care as a key area of opportunity for improving postdischarge care4

Lack of communication has clearly been shown to adversely affect postdischarge care transitions.5 A recent summary of the literature by a Society of Hospital Medicine (SHM)/Society of General Internal Medicine (SGIM) task force found that direct communication between hospital physicians and primary care physicians occurs infrequently (in 3%‐20% of cases studied), and the availability of a discharge summary at the first postdischarge visit is low (12%‐34%) and does not improve greatly even after 4 weeks (51%‐77%); this affects the quality of care in approximately 25% of follow‐up visits.5 This systematic review of the literature also found that discharge summaries often lack important information such as diagnostic test results, the treatment or hospital course, discharge medications, test results pending at discharge, patient or family counseling, and follow‐up plans.

However, the lack of studies of the communication between ambulatory physicians and hospital physicians prior to admission or during emergency department (ED) visits does not imply that this communication is not equally important and essential to high‐quality care. According to the Centers for Disease Control, the greatest source of hospital admissions in many institutions is the ED. Over 115,000,000 visits were made to the nation's approximately 4828 EDs in 2005, and about 85.2% of ED visits end in discharge.6 The ED is also the point of re‐entry into the system for individuals who may have had an adverse outcome linked to a prior hospitalization.6 Communication between hospital physicians and primary care physicians must be established to create a loop of continuous care and diminish morbidity and mortality at this critical transition point.

While transitions can be a risky period for patient safety, observational studies suggest there are benefits to transitions. A new physician may notice something overlooked by the current caregivers.712 Another factor contributing to the challenges of care transitions is the lack of a single clinician or clinical entity taking responsibility for coordination across the continuum of the patient's overall healthcare, regardless of setting.13 Studies indicate that a relationship with a medical home is associated with better health on both the individual and population levels, with lower overall costs of care and with reductions in disparities in health between socially disadvantaged subpopulations and more socially advantaged populations.14 Several medical societies have addressed this issue, including the American College of Physicians (ACP), SGIM, American Academy of Family Physicians, and American Academy of Pediatrics, and they have proposed the concept of the medical home or patient‐centered medical home, which calls for clinicians to assume this responsibility for coordinating their patients' care across settings and for the healthcare system to value and reimburse clinicians for this patient‐centered and comprehensive method of practice.1517

Finally, patients and their families or caregivers have an important role to play in transitions of care. Several observational and cross‐sectional studies have shown that patients and their caregivers and families express significant feelings of anxiety during care transitions. This anxiety can be caused by a lack of understanding and preparation for their self‐care role in the next care setting, confusion due to conflicting advice from different practitioners, and a sense of abandonment attributable to the inability to contact an appropriate healthcare practitioner for guidance, and they report an overall disregard for their preferences and input into the design of the care plan.1820 Clearly, there is room for improvement in all these areas of the inpatient and outpatient care transition, and the Transitions of Care Consensus Conference (TOCCC) attempted to address these areas by developing standards for the transition of care that also harmonize with the work of the Stepping up to the Plate (SUTTP) Alliance of the American Board of Internal Medicine (ABIM) Foundation.21 In addition, other important stakeholders are addressing this topic and actively working to improve communication and continuity in care, including the Centers for Medicare and Medicaid Services (CMS) and the National Quality Forum (NQF). CMS recently developed the Continuity Assessment Record & Evaluation (CARE) tool, a data collection instrument designed to be a standardized, interoperable, common assessment tool to capture key patient characteristics that will provide information related to resource utilization, clinical outcomes, and postdischarge disposition. NQF held a national forum on care coordination in the spring of 2008.

In summary, it is clear that there are qualitative and quantitative deficiencies in transitions of care between the inpatient and outpatient setting that are affecting patient safety and experience with care. The transition from the inpatient setting to the outpatient setting has been more extensively studied, and this body of literature has underscored for the TOCCC several important areas in need of guidance and improvement. Because of this, the scope of application of this document should initially emphasize inpatient‐to‐outpatient transitions as a first step in learning how to improve these processes. However, the transition from the outpatient setting to the inpatient setting also is a clear priority. Because the needs for transfer of information, authority, and responsibility may be different in these situations, a second phase of additional work to develop principles to guide these transitions should be undertaken as quickly as possible. Experience gained in applying these principles to inpatient‐to‐outpatient transitions might usefully inform such work.

Communication among providers and with the patients and their families arose as a clear priority. Medication discrepancies, pending tests, and unknown diagnostic or treatment plans have an immediate impact on patients' health and outcomes. The TOCCC discussed what elements should be among the standard pieces of information exchanged among providers during these transition points. The dire need for coordination of care or a coordinating clinician/medical home became a clear theme in the deliberations of the TOCCC. Most importantly, the role of the patients and their families/caregivers in their continuing care is apparent, and the TOCCC felt this must be an integral part of any principles or standards for transitions of care.

Methods

In the fall/winter of 2006, the executive committees of ACP, SGIM, and SHM agreed to jointly develop a policy statement on transitions of care. Transitions of care specifically between the inpatient and outpatient settings were selected as an ideal topic for collaboration for the 3 societies as they represent the continuum of care for internal medicine within these settings. To accomplish this, the 3 organizations decided to convene a consensus conference to develop consensus guidelines and standards concerning transitions between inpatient and outpatient settings through a multi‐stakeholder process. A steering committee was convened with representatives from ACP, SGIM, SHM, the Agency for Healthcare Research and Quality (AHRQ), ABIM, and the American Geriatric Society (AGS). The steering committee developed the agenda and invitee list for the consensus conference. After the conference was held, the steering committee was expanded to include representation from the American College of Emergency Physicians (ACEP) and the Society for Academic Emergency Medicine (SAEM).

During the planning stages of the TOCCC, the steering committee became aware of the SUTTP Alliance of the ABIM Foundation. The SUTTP Alliance has representation from medical specialties such as internal medicine and its subspecialties, family medicine, and surgery. The alliance was formed in 2006 and has been working on care coordination across multiple settings and specialties. The SUTTP Alliance had developed a set of principles and standards for care transitions and agreed to provide the draft document to the TOCCC for review, input, and further development and refinement.

Recommendations on Principles and Standards for Managing Transitions in Care Between the Inpatient and Outpatient Settings from ACP, SGIM, SHM, AGS, ACEP, and SAEM

The SUTTP Alliance presented a draft document entitled Principles and Standards for Managing Transitions in Care. In this document, the SUTTP Alliance proposes 5 principles and 8 standards for effective care transitions. A key element of the conference was a presentation by NQF on how to move from principles to standards and eventually to measures. This presentation provided the TOCCC with the theoretical underpinnings for the discussion of these principles and standards and how the TOCCC would provide input on them. The presentation provided an outline for the flow from principles to measures. First, there needs to be a framework that provides guiding principles for what we would like to measure and eventually report. From those principles, a set of preferred practices or standards are developed; the standards are more granular and allow for more specificity in describing the desired practice or outcome and its elements. Standards then provide a roadmap for identification and development of performance measures. With this framework in mind, the TOCCC then discussed in detail the SUTTP principles and standards.

The 5 principles for effective care transitions developed by the SUTTP Alliance are as follows:

• Accountability.

• Communication: clear and direct communication of treatment plans and follow‐up expectations.

• Timely feedback and feed‐forward of information.

• Involvement of the patient and family member, unless inappropriate, in all steps.

• Respect of the hub of coordination of care.

The TOCCC re‐affirmed these principles and added 4 additional principles to this list. Three of the new principles were statements within the 8 standards developed by the SUTTP, but when taking into consideration the framework for the development of principles into standards, the TOCCC felt that the statements were better represented as principles. They are as follows:

• All patients and their families/caregivers should have and should be able to identify their medical home or coordinating clinician (ie, practice or practitioner). (This was originally part of the coordinating clinicians standard, and the TOCCC voted to elevate this to a principle).

• At every point along the transition, the patients and/or their families/caregivers need to know who is responsible for care at that point and who to contact and how.

• National standards should be established for transitions in care and should be adopted and implemented at the national and community level through public health institutions, national accreditation bodies, medical societies, medical institutions, and so forth in order to improve patient outcomes and patient safety. (This was originally part of the SUTTP community standards standard, and the TOCCC moved to elevate this to a principle).

• For monitoring and improving transitions, standardized metrics related to these standards should be used in order to lead to continuous quality improvement and accountability. (This was originally part of the measurement standard, and the TOCCC voted to elevate this to a principle).

The SUTTP Alliance proposed the following 8 standards for care transitions:

• Coordinating clinicians.

• Care plans.

• Communication infrastructure.

• Standard communication formats.

• Transition responsibility.

• Timeliness.

• Community standards.

• Measurement.

The TOCCC affirmed these standards and through a consensus process added more specificity to most of them and elevated components of some of them to principles, as discussed previously. The TOCCC proposes that the following be merged with the SUTTP standards:

• Coordinating clinicians. Communication and information exchange between the medical home and the receiving provider should occur in an amount of time that will allow the receiving provider to effectively treat the patient. This communication and information exchange should ideally occur whenever patients are at a transition of care (eg, at discharge from the inpatient setting). The timeliness of this communication should be consistent with the patient's clinical presentation and, in the case of a patient being discharged, the urgency of the follow‐up required. Guidelines will need to be developed that address both the timeliness and means of communication between the discharging physician and the medical home. Communication and information exchange between the medical home and other physicians may be in the form of a call, voice mail, fax, or other secure, private, and accessible means including mutual access to an electronic health record.

  The ED represents a unique subset of transitions of care. The potential transition can generally be described as outpatient to outpatient or outpatient to inpatient, depending on whether or not the patient is admitted to the hospital. The outpatient‐to‐outpatient transition can also encompass a number of potential variations. Patients with a medical home may be referred to the ED by the medical home, or they may self‐refer. A significant number of patients do not have a physician and self‐refer to the ED. The disposition from the ED, either outpatient to outpatient or outpatient to inpatient, is similarly represented by a number of variables. Discharged patients may or may not have a medical home, may or may not need a specialist, and may or may not require urgent (<24 hours) follow‐up. Admitted patients may or may not have a medical home and may or may not require specialty care. This variety of variables precludes a single approach to ED transition of care coordination. The determination of which scenarios will be appropriate for the development of standards (coordinating clinicians and transition responsibility) will require further contributions from ACEP and SAEM and review by the steering committee.

• Care plans/transition record. The TOCCC also agreed that there is a minimal set of data elements that should always be part of the transition record. The TOCCC suggested that this minimal data set be part of an initial implementation of this standard. That list includes the following:

  The TOCCC discussed what components should be included in an ideal transition record and agreed on the following elements:

  The TOCCC also added a new standard under this heading: Patients and/or their families/caregivers must receive, understand, and be encouraged to participate in the development of the transition record, which should take into consideration patients' health literacy and insurance status and be culturally sensitive.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Clear identification of the medical home/transferring coordinating physician/emnstitution and the contact information.

• Patient's cognitive status.

• Test results/pending results.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Emergency plan and contact number and person.

• Treatment and diagnostic plan.

• Prognosis and goals of care.

• Test results/pending results.

• Clear identification of the medical home and/or transferring coordinating physician/emnstitution.

• Patient's cognitive status.

• Advance directives, power of attorney, and consent.

• Planned interventions, durable medical equipment, wound care, and so forth.

• Assessment of caregiver status.

• Communication infrastructure. All communications between providers and between providers and patients and families/caregivers need to be secure, private, Health Insurance Portability and Accountability Actcompliant, and accessible to patients and those practitioners who care for them. Communication needs to be 2‐way with an opportunity for clarification and feedback. Each sending provider needs to provide a contact name and the number of an individual who can respond to questions or concerns. The content of transferred information needs to include a core standardized data set. This information needs to be transferred as a living database; that is, it is created only once, and then each subsequent provider only needs to update, validate, or modify the information. Patient information should be available to the provider prior to the patient's arrival. Information transfer needs to adhere to national data standards. Patients should be provided with a medication list that is accessible (paper or electronic), clear, and dated.

• Standard communication formats. Communities need to develop standard data transfer forms (templates and transmission protocols). Access to a patient's medical history needs to be on a current and ongoing basis with the ability to modify information as a patient's condition changes. Patients, families, and caregivers should have access to their information (nothing about me without me). A section on the transfer record should be devoted to communicating a patient's preferences, priorities, goals, and values (eg, the patient does not want intubation).

• Transition responsibility. The sending provider/emnstitution/team at the clinical organization maintains responsibility for the care of the patient until the receiving clinician/location confirms that the transfer and assumption of responsibility is complete (within a reasonable timeframe for the receiving clinician to receive the information; ie, transfers that occur in the middle of the night can be communicated during standard working hours). The sending provider should be available for clarification with issues of care within a reasonable timeframe after the transfer has been completed, and this timeframe should be based on the conditions of the transfer settings. The patient should be able to identify the responsible provider. In the case of patients who do not have an ongoing ambulatory care provider or whose ambulatory care provider has not assumed responsibility, the hospital‐based clinicians will not be required to assume responsibility for the care of these patients once they are discharged.

• Timeliness. Timeliness of feedback and feed‐forward of information from a sending provider to a receiving provider should be contingent on 4 factors:

  This information should be available at the time of the patient encounter.

• Transition settings.

• Patient circumstances.

• Level of acuity.

• Clear transition responsibility.

• Community standards. Medical communities/emnstitutions must demonstrate accountability for transitions of care by adopting national standards, and processes should be established to promote effective transitions of care.

• Measurement. For monitoring and improving transitions, standardized metrics related to these standards should be used. These metrics/measures should be evidence‐based, address documented gaps, and have a demonstrated impact on improving care (complying with performance measure standards) whenever feasible. Results from measurements using standardized metrics must lead to continuous improvement of the transition process. The validity, reliability, cost, and impact, including unintended consequences, of these measures should be assessed and re‐evaluated.

All these standards should be applied with special attention to the various transition settings and should be appropriate to each transition setting. Measure developers will need to take this into account when developing measures based on these proposed standards.

The TOCCC also went through a consensus prioritization exercise to rank‐order the consensus standards. All meeting participants were asked to rank their top 3 priorities of the 7 standards, giving a numeric score of 1 for their highest priority, a score of 2 for their second highest priority, and a score of 3 for their third highest priority. Summary scores were calculated, and the standards were rank‐ordered from the lowest summary score to the highest. The TOCCC recognizes that full implementation of all of these standards may not be feasible and that these standards may be implemented on a stepped or incremental basis. This prioritization can assist in deciding which of these to implement. The results of the prioritization exercise are as follows:

• 1

  All transitions must include a transition record

• 2

  Transition responsibility

• 3

  Coordinating clinicians

• 4

  Patient and family involvement and ownership of the transition record

• 5

  Communication infrastructure

• 6

  Timeliness

• 7

  Community standards

Future Challenges

In addition to the work on the principles and standards, the TOCCC uncovered six further challenges which are described below.

Next Steps for the TOCCC

The TOCCC focuses only on the transitions between the inpatient and outpatient settings and does not address the equally important transitions between many other different care settings, such as the transition from a hospital to a nursing home or rehabilitation facility. The intent of the TOCCC is to provide this document to national measure developers such as the Physician Consortium for Performance Improvement and others in order to guide measure development and ultimately lead to improvements in quality and safety in care transitions.

Studies of the transition of care between inpatient and outpatient settings have shown that there are significant patient safety and quality deficiencies in our current system. The transition from the hospital setting to the outpatient setting has been more extensively studied than the transition from the outpatient setting to the inpatient setting. One prospective cohort study of 400 patients found that 1 in 5 patients discharged from the hospital to home experienced an adverse event, which was defined as an injury resulting from medical management rather than the underlying disease, within 3 weeks of discharge.1 This study also concluded that 66% of these were drug‐related adverse events, many of which could have been avoided or mitigated. Another prospective cross‐sectional study of 2644 patient discharges found that approximately 40% of the patients had pending test results at the time of discharge and that 10% of these required some action, yet the outpatient physicians and patients were unaware of these results.2 Medication discrepancies have also been shown to be prevalent, with 1 prospective observational study of 375 patients showing that 14% of elderly patients had 1 or more medication discrepancies and 14% of those patients with medication discrepancies were rehospitalized within 30 days versus 6% of the patients who did not experience a medication discrepancy.3 A recent review of the literature cited improving transitional care as a key area of opportunity for improving postdischarge care4

Lack of communication has clearly been shown to adversely affect postdischarge care transitions.5 A recent summary of the literature by a Society of Hospital Medicine (SHM)/Society of General Internal Medicine (SGIM) task force found that direct communication between hospital physicians and primary care physicians occurs infrequently (in 3%‐20% of cases studied), and the availability of a discharge summary at the first postdischarge visit is low (12%‐34%) and does not improve greatly even after 4 weeks (51%‐77%); this affects the quality of care in approximately 25% of follow‐up visits.5 This systematic review of the literature also found that discharge summaries often lack important information such as diagnostic test results, the treatment or hospital course, discharge medications, test results pending at discharge, patient or family counseling, and follow‐up plans.

However, the lack of studies of the communication between ambulatory physicians and hospital physicians prior to admission or during emergency department (ED) visits does not imply that this communication is not equally important and essential to high‐quality care. According to the Centers for Disease Control, the greatest source of hospital admissions in many institutions is the ED. Over 115,000,000 visits were made to the nation's approximately 4828 EDs in 2005, and about 85.2% of ED visits end in discharge.6 The ED is also the point of re‐entry into the system for individuals who may have had an adverse outcome linked to a prior hospitalization.6 Communication between hospital physicians and primary care physicians must be established to create a loop of continuous care and diminish morbidity and mortality at this critical transition point.

While transitions can be a risky period for patient safety, observational studies suggest there are benefits to transitions. A new physician may notice something overlooked by the current caregivers.712 Another factor contributing to the challenges of care transitions is the lack of a single clinician or clinical entity taking responsibility for coordination across the continuum of the patient's overall healthcare, regardless of setting.13 Studies indicate that a relationship with a medical home is associated with better health on both the individual and population levels, with lower overall costs of care and with reductions in disparities in health between socially disadvantaged subpopulations and more socially advantaged populations.14 Several medical societies have addressed this issue, including the American College of Physicians (ACP), SGIM, American Academy of Family Physicians, and American Academy of Pediatrics, and they have proposed the concept of the medical home or patient‐centered medical home, which calls for clinicians to assume this responsibility for coordinating their patients' care across settings and for the healthcare system to value and reimburse clinicians for this patient‐centered and comprehensive method of practice.1517

Finally, patients and their families or caregivers have an important role to play in transitions of care. Several observational and cross‐sectional studies have shown that patients and their caregivers and families express significant feelings of anxiety during care transitions. This anxiety can be caused by a lack of understanding and preparation for their self‐care role in the next care setting, confusion due to conflicting advice from different practitioners, and a sense of abandonment attributable to the inability to contact an appropriate healthcare practitioner for guidance, and they report an overall disregard for their preferences and input into the design of the care plan.1820 Clearly, there is room for improvement in all these areas of the inpatient and outpatient care transition, and the Transitions of Care Consensus Conference (TOCCC) attempted to address these areas by developing standards for the transition of care that also harmonize with the work of the Stepping up to the Plate (SUTTP) Alliance of the American Board of Internal Medicine (ABIM) Foundation.21 In addition, other important stakeholders are addressing this topic and actively working to improve communication and continuity in care, including the Centers for Medicare and Medicaid Services (CMS) and the National Quality Forum (NQF). CMS recently developed the Continuity Assessment Record & Evaluation (CARE) tool, a data collection instrument designed to be a standardized, interoperable, common assessment tool to capture key patient characteristics that will provide information related to resource utilization, clinical outcomes, and postdischarge disposition. NQF held a national forum on care coordination in the spring of 2008.

In summary, it is clear that there are qualitative and quantitative deficiencies in transitions of care between the inpatient and outpatient setting that are affecting patient safety and experience with care. The transition from the inpatient setting to the outpatient setting has been more extensively studied, and this body of literature has underscored for the TOCCC several important areas in need of guidance and improvement. Because of this, the scope of application of this document should initially emphasize inpatient‐to‐outpatient transitions as a first step in learning how to improve these processes. However, the transition from the outpatient setting to the inpatient setting also is a clear priority. Because the needs for transfer of information, authority, and responsibility may be different in these situations, a second phase of additional work to develop principles to guide these transitions should be undertaken as quickly as possible. Experience gained in applying these principles to inpatient‐to‐outpatient transitions might usefully inform such work.

Communication among providers and with the patients and their families arose as a clear priority. Medication discrepancies, pending tests, and unknown diagnostic or treatment plans have an immediate impact on patients' health and outcomes. The TOCCC discussed what elements should be among the standard pieces of information exchanged among providers during these transition points. The dire need for coordination of care or a coordinating clinician/medical home became a clear theme in the deliberations of the TOCCC. Most importantly, the role of the patients and their families/caregivers in their continuing care is apparent, and the TOCCC felt this must be an integral part of any principles or standards for transitions of care.

Methods

In the fall/winter of 2006, the executive committees of ACP, SGIM, and SHM agreed to jointly develop a policy statement on transitions of care. Transitions of care specifically between the inpatient and outpatient settings were selected as an ideal topic for collaboration for the 3 societies as they represent the continuum of care for internal medicine within these settings. To accomplish this, the 3 organizations decided to convene a consensus conference to develop consensus guidelines and standards concerning transitions between inpatient and outpatient settings through a multi‐stakeholder process. A steering committee was convened with representatives from ACP, SGIM, SHM, the Agency for Healthcare Research and Quality (AHRQ), ABIM, and the American Geriatric Society (AGS). The steering committee developed the agenda and invitee list for the consensus conference. After the conference was held, the steering committee was expanded to include representation from the American College of Emergency Physicians (ACEP) and the Society for Academic Emergency Medicine (SAEM).

During the planning stages of the TOCCC, the steering committee became aware of the SUTTP Alliance of the ABIM Foundation. The SUTTP Alliance has representation from medical specialties such as internal medicine and its subspecialties, family medicine, and surgery. The alliance was formed in 2006 and has been working on care coordination across multiple settings and specialties. The SUTTP Alliance had developed a set of principles and standards for care transitions and agreed to provide the draft document to the TOCCC for review, input, and further development and refinement.

Recommendations on Principles and Standards for Managing Transitions in Care Between the Inpatient and Outpatient Settings from ACP, SGIM, SHM, AGS, ACEP, and SAEM

The SUTTP Alliance presented a draft document entitled Principles and Standards for Managing Transitions in Care. In this document, the SUTTP Alliance proposes 5 principles and 8 standards for effective care transitions. A key element of the conference was a presentation by NQF on how to move from principles to standards and eventually to measures. This presentation provided the TOCCC with the theoretical underpinnings for the discussion of these principles and standards and how the TOCCC would provide input on them. The presentation provided an outline for the flow from principles to measures. First, there needs to be a framework that provides guiding principles for what we would like to measure and eventually report. From those principles, a set of preferred practices or standards are developed; the standards are more granular and allow for more specificity in describing the desired practice or outcome and its elements. Standards then provide a roadmap for identification and development of performance measures. With this framework in mind, the TOCCC then discussed in detail the SUTTP principles and standards.

The 5 principles for effective care transitions developed by the SUTTP Alliance are as follows:

• Accountability.

• Communication: clear and direct communication of treatment plans and follow‐up expectations.

• Timely feedback and feed‐forward of information.

• Involvement of the patient and family member, unless inappropriate, in all steps.

• Respect of the hub of coordination of care.

The TOCCC re‐affirmed these principles and added 4 additional principles to this list. Three of the new principles were statements within the 8 standards developed by the SUTTP, but when taking into consideration the framework for the development of principles into standards, the TOCCC felt that the statements were better represented as principles. They are as follows:

• All patients and their families/caregivers should have and should be able to identify their medical home or coordinating clinician (ie, practice or practitioner). (This was originally part of the coordinating clinicians standard, and the TOCCC voted to elevate this to a principle).

• At every point along the transition, the patients and/or their families/caregivers need to know who is responsible for care at that point and who to contact and how.

• National standards should be established for transitions in care and should be adopted and implemented at the national and community level through public health institutions, national accreditation bodies, medical societies, medical institutions, and so forth in order to improve patient outcomes and patient safety. (This was originally part of the SUTTP community standards standard, and the TOCCC moved to elevate this to a principle).

• For monitoring and improving transitions, standardized metrics related to these standards should be used in order to lead to continuous quality improvement and accountability. (This was originally part of the measurement standard, and the TOCCC voted to elevate this to a principle).

The SUTTP Alliance proposed the following 8 standards for care transitions:

• Coordinating clinicians.

• Care plans.

• Communication infrastructure.

• Standard communication formats.

• Transition responsibility.

• Timeliness.

• Community standards.

• Measurement.

The TOCCC affirmed these standards and through a consensus process added more specificity to most of them and elevated components of some of them to principles, as discussed previously. The TOCCC proposes that the following be merged with the SUTTP standards:

• Coordinating clinicians. Communication and information exchange between the medical home and the receiving provider should occur in an amount of time that will allow the receiving provider to effectively treat the patient. This communication and information exchange should ideally occur whenever patients are at a transition of care (eg, at discharge from the inpatient setting). The timeliness of this communication should be consistent with the patient's clinical presentation and, in the case of a patient being discharged, the urgency of the follow‐up required. Guidelines will need to be developed that address both the timeliness and means of communication between the discharging physician and the medical home. Communication and information exchange between the medical home and other physicians may be in the form of a call, voice mail, fax, or other secure, private, and accessible means including mutual access to an electronic health record.

  The ED represents a unique subset of transitions of care. The potential transition can generally be described as outpatient to outpatient or outpatient to inpatient, depending on whether or not the patient is admitted to the hospital. The outpatient‐to‐outpatient transition can also encompass a number of potential variations. Patients with a medical home may be referred to the ED by the medical home, or they may self‐refer. A significant number of patients do not have a physician and self‐refer to the ED. The disposition from the ED, either outpatient to outpatient or outpatient to inpatient, is similarly represented by a number of variables. Discharged patients may or may not have a medical home, may or may not need a specialist, and may or may not require urgent (<24 hours) follow‐up. Admitted patients may or may not have a medical home and may or may not require specialty care. This variety of variables precludes a single approach to ED transition of care coordination. The determination of which scenarios will be appropriate for the development of standards (coordinating clinicians and transition responsibility) will require further contributions from ACEP and SAEM and review by the steering committee.

• Care plans/transition record. The TOCCC also agreed that there is a minimal set of data elements that should always be part of the transition record. The TOCCC suggested that this minimal data set be part of an initial implementation of this standard. That list includes the following:

  The TOCCC discussed what components should be included in an ideal transition record and agreed on the following elements:

  The TOCCC also added a new standard under this heading: Patients and/or their families/caregivers must receive, understand, and be encouraged to participate in the development of the transition record, which should take into consideration patients' health literacy and insurance status and be culturally sensitive.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Clear identification of the medical home/transferring coordinating physician/emnstitution and the contact information.

• Patient's cognitive status.

• Test results/pending results.

• Principle diagnosis and problem list.

• Medication list (reconciliation) including over‐the‐counter medications/herbals, allergies, and drug interactions.

• Emergency plan and contact number and person.

• Treatment and diagnostic plan.

• Prognosis and goals of care.

• Test results/pending results.

• Clear identification of the medical home and/or transferring coordinating physician/emnstitution.

• Patient's cognitive status.

• Advance directives, power of attorney, and consent.

• Planned interventions, durable medical equipment, wound care, and so forth.

• Assessment of caregiver status.

• Communication infrastructure. All communications between providers and between providers and patients and families/caregivers need to be secure, private, Health Insurance Portability and Accountability Actcompliant, and accessible to patients and those practitioners who care for them. Communication needs to be 2‐way with an opportunity for clarification and feedback. Each sending provider needs to provide a contact name and the number of an individual who can respond to questions or concerns. The content of transferred information needs to include a core standardized data set. This information needs to be transferred as a living database; that is, it is created only once, and then each subsequent provider only needs to update, validate, or modify the information. Patient information should be available to the provider prior to the patient's arrival. Information transfer needs to adhere to national data standards. Patients should be provided with a medication list that is accessible (paper or electronic), clear, and dated.

• Standard communication formats. Communities need to develop standard data transfer forms (templates and transmission protocols). Access to a patient's medical history needs to be on a current and ongoing basis with the ability to modify information as a patient's condition changes. Patients, families, and caregivers should have access to their information (nothing about me without me). A section on the transfer record should be devoted to communicating a patient's preferences, priorities, goals, and values (eg, the patient does not want intubation).

• Transition responsibility. The sending provider/emnstitution/team at the clinical organization maintains responsibility for the care of the patient until the receiving clinician/location confirms that the transfer and assumption of responsibility is complete (within a reasonable timeframe for the receiving clinician to receive the information; ie, transfers that occur in the middle of the night can be communicated during standard working hours). The sending provider should be available for clarification with issues of care within a reasonable timeframe after the transfer has been completed, and this timeframe should be based on the conditions of the transfer settings. The patient should be able to identify the responsible provider. In the case of patients who do not have an ongoing ambulatory care provider or whose ambulatory care provider has not assumed responsibility, the hospital‐based clinicians will not be required to assume responsibility for the care of these patients once they are discharged.

• Timeliness. Timeliness of feedback and feed‐forward of information from a sending provider to a receiving provider should be contingent on 4 factors:

  This information should be available at the time of the patient encounter.

• Transition settings.

• Patient circumstances.

• Level of acuity.

• Clear transition responsibility.

• Community standards. Medical communities/emnstitutions must demonstrate accountability for transitions of care by adopting national standards, and processes should be established to promote effective transitions of care.

• Measurement. For monitoring and improving transitions, standardized metrics related to these standards should be used. These metrics/measures should be evidence‐based, address documented gaps, and have a demonstrated impact on improving care (complying with performance measure standards) whenever feasible. Results from measurements using standardized metrics must lead to continuous improvement of the transition process. The validity, reliability, cost, and impact, including unintended consequences, of these measures should be assessed and re‐evaluated.

All these standards should be applied with special attention to the various transition settings and should be appropriate to each transition setting. Measure developers will need to take this into account when developing measures based on these proposed standards.

The TOCCC also went through a consensus prioritization exercise to rank‐order the consensus standards. All meeting participants were asked to rank their top 3 priorities of the 7 standards, giving a numeric score of 1 for their highest priority, a score of 2 for their second highest priority, and a score of 3 for their third highest priority. Summary scores were calculated, and the standards were rank‐ordered from the lowest summary score to the highest. The TOCCC recognizes that full implementation of all of these standards may not be feasible and that these standards may be implemented on a stepped or incremental basis. This prioritization can assist in deciding which of these to implement. The results of the prioritization exercise are as follows:

• 1

  All transitions must include a transition record

• 2

  Transition responsibility

• 3

  Coordinating clinicians

• 4

  Patient and family involvement and ownership of the transition record

• 5

  Communication infrastructure

• 6

  Timeliness

• 7

  Community standards

Future Challenges

In addition to the work on the principles and standards, the TOCCC uncovered six further challenges which are described below.

Next Steps for the TOCCC

The TOCCC focuses only on the transitions between the inpatient and outpatient settings and does not address the equally important transitions between many other different care settings, such as the transition from a hospital to a nursing home or rehabilitation facility. The intent of the TOCCC is to provide this document to national measure developers such as the Physician Consortium for Performance Improvement and others in order to guide measure development and ultimately lead to improvements in quality and safety in care transitions.

A 71‐year‐old African‐American woman presented to the emergency department with chest pain, shortness of breath, and cough. She had initially presented to her primary care physician 2 weeks previously complaining of worsening cough and shortness of breath and was told to continue her inhaled albuterol and glucocorticoids and was prescribed a prednisone taper and an unknown course of antibiotics. She noted no improvement in her symptoms despite compliance with this treatment. Three days prior to admission she described the gradual onset of left‐sided pleuritic chest pain with continued cough, associated with yellow sputum and worsening dyspnea. Review of systems was remarkable for generalized weakness and malaise. She denied fever, chills, orthopnea, paroxysmal nocturnal dyspnea, lower extremity edema, diarrhea, nausea, vomiting, or abdominal pain.

Her past medical history included a diagnosis of chronic obstructive pulmonary disease (COPD) but pulmonary function tests 7 years prior to admission showed an forced expiratory volume in the first second (FEV1)/forced vital capacity (FVC) ratio of 81%. She had a 30 pack‐year history of smoking, but quit 35 years ago. The patient also carried a diagnosis of heart failure, but an echocardiogram done 1 year ago demonstrated a left ventricular ejection fraction of 65% to 70% without diastolic dysfunction but mild right ventricular dilation and hypertrophy. Additionally, she had known nonobstructive coronary atherosclerotic heart disease, dyslipidemia, hypertension, morbid obesity, depression, and a documented chronic right hemidiaphragm elevation.

At this point the history suggests that the patient does not have a clear diagnosis of COPD. The lack of definitive spirometry evidence of chronic airway obstruction concerns me; I think that she may have been mistakenly treated with chronic inhaled steroids and doses of antibiotics for an acute exacerbation of chronic lung disease. Additional review of her history gives some indication of advanced lung disease, with her recent echocardiogram showing strain on the right ventricle with right ventricular hypertrophy and dilation, but there is no mention of the presence or severity of pulmonary hypertension. Nonetheless, I would be concerned that she probably has underlying significant cor pulmonale.

The patient now re‐presents with a worsening of her pulmonary symptoms. Her left‐sided pleuritic pain would make me concerned that she had a pulmonary embolus (PE). This morbidly obese patient with new pulmonary symptoms, right ventricular strain on her previous echocardiogram, and a persistent elevated right hemidiaphragm suggests a presentation of another PE.

At this time I cannot rule out other common possibilities such as infectious pneumonia. If she does have pneumonia, I would be concerned she could be harboring a multidrug‐resistant bacterial infection given her recent course of antibiotics in addition to her use of both chronic inhaled and intermittent oral glucocorticoids.

After gathering the rest of her full medical history, I would focus my physical exam on looking for evidence of parenchymal lung disease, signs of pulmonary hypertension, and pneumonia.

Her surgical history includes a previous hysterectomy, cholecystectomy, hernia repair, and left hepatic lobectomy for a benign mass. Her outpatient medications were ibuprofen, bupropion, fluvastatin, atenolol, potassium, aspirin, clopidogrel, albuterol inhaler, fluticasone/salmeterol inhaler, and omeprazole. She reports an allergy to penicillin and to sulfa drugs. Her mother died of an unknown cancer at age 77 years. She denied any international travel and she has always lived in Georgia.

The patient has been retired since 1992, having previously worked for the U.S. Postal Service. She admits to occasional alcohol intake (2 to 3 drinks a month). No recent travel, surgery, or prolonged immobilization was noted.

On initial examination she was alert and mentally appropriate, but appeared to be in mild respiratory distress with a respiratory rate of 28 breaths/minute. Her blood pressure (BP) was 99/70, heart rate 102, temperature of 38.2C, and oxygen saturation of 93% on room air and 97% on 2 L of oxygen via nasal cannula. Auscultation of her lungs revealed crackles over her left anterior lung field, bronchial breath sounds in the left posterior midlung, and bibasilar crackles. No wheezing was noted. Her cardiovascular exam and the remainder of her physical exam were unremarkable except for morbid obesity.

While my initial thoughts were leaning toward an exacerbation of chronic lung disease or possibly a new PE, at this moment, infection seems more likely. Indeed, her pulmonary findings suggest a left‐sided inflammatory process, and her vital signs meet criteria for systemic inflammatory response syndrome (SIRS). My primary concern is sepsis due to a drug‐resistant bacterial infection, including Staphylococcus aureus or gram‐negative bacteria or possibly more unusual organisms such as Nocardia or fungi, due to her recent use of antibiotics and chronic inhaled steroid use and recent course of oral glucocorticoids.

Conversely, the SIRS could be a manifestation of a noninfectious lung process such as acute interstitial pneumonia or an eosinophilic pneumonia. Given the diagnostic complexity, I would strongly consider consulting a pulmonologist if the patient did not improve quickly. At this point, I would like to review a posterior‐anterior (PA) and lateral chest radiograph, and room air arterial blood gas (ABG) in addition to basic laboratory test values.

Laboratory data obtained on admission was remarkable for a white blood cell (WBC) count of 26,500/L with 75% neutrophils and 6% eosinophils. Hemoglobin was 14.4 gm/dL. Platelet count was 454,000/L. Serum chemistries showed a sodium of 137 mEq/dL, potassium 4.3 mEq/dL, Cl 108 mEq/dL, bicarbonate 19 mEq/dL, blood urea nitrogen (BUN) 8 mg/dL, creatinine 1.0 mg/dL, and glucose 137 mg/dL. Cardiac enzymes were normal. Calcium was 9.8 mg/dL, albumin 2.7 gm/dL, total protein 6.9 gm/dL, AST 36 U/L, ALT 54 U/L and the bilirubin was normal. Chest radiograph (Figure 1) demonstrated a left perihilar infiltrate with air bronchograms and marked right hemidiaphragm elevation as seen on previous films. Unchanged increased interstitial markings were also present. Her electrocardiogram (ECG) showed normal sinus rhythm, normal axis, and QRS duration with nonspecific diffuse T‐wave abnormalities.

Figure 1

Given her presentation, I am worried about how well she is oxygenating and ventilating. An ABG should be done to assess her status more accurately. An albumin of 2.7 gm/dL indicates that she is fairly sick. I would not hesitate to consider testing the patient for human immunodeficiency virus (HIV) given how this information would dramatically change the differential diagnoses of her pulmonary process.

I am still most concerned about sepsis secondary to pneumonia in this patient with multiple chronic comorbidities, underlying chronic lung disease, receiving chronic inhaled glucocorticoids and a recent course of oral glucocorticoids and antibiotics. While I would initiate hydration I do not see a clear indication for early goal‐directed therapy for severe sepsis. In addition to obtaining an ABG and starting intravenous fluids, I would also draw blood cultures, send sputum for gram stain, culture, and sensitivity, and perform a urinalysis. I would also administer empiric antibiotics as quickly as possible based on a number of pneumonia clinical studies suggesting improved outcomes with early antibiotic administration. Because of her use of antibiotics and both inhaled and oral glucocorticoids, she is at higher risk for potentially multidrug‐resistant bacterial pathogens, including Staphyloccocus aureus and gram‐negative bacteria such as Pseudomonas and Klebsiella (Table 1). Therefore, I would initially cover her broadly for these organisms.

In addition to initial treatment choice, the inpatient triage decision is another important issue, especially at a community hospital where intensive care unit (ICU) resources are rare and often the admission decision is between sending a moderately sick patient to a regular floor bed or the medical ICU. Both the American Thoracic Society and Infectious Diseases Society of America support an ICU triage protocol in their guidelines for the management of community‐acquired pneumonia in adults that utilizes the following 9 minor criteria, of which the presence of at least 3 should support ICU admission: respiratory rate 30 breaths/minute; oxygenation index (pressure of oxygen [PaO₂]/fraction of inspired oxygen [FiO₂] ratio) 250; multilobar infiltrates; confusion/disorientation; uremia (BUN level 20 mg/dL); leukopenia (WBC count <4,000 cells/mm³); thrombocytopenia (platelet count <100,000 cells/mm³); hypothermia (core temperature <36C); and hypotension requiring aggressive fluid. Despite the absence of these criteria in this patient, it is important to note that no triage protocol has been adequately prospectively validated. Retrospective study of the minor criteria has found that the presence of at least 2 of the following 3 clinical criteria to have the highest specificity for predicting cardiopulmonary decompensation and subsequent need for ICU care: (1) initial hypotension (BP <90/60) on presentation with response to initial intravenous fluids to a BP >90/60; (2) oxygenation failure as indicated by PaO₂/FiO₂ ratio less than 250; or (3) the presence of multilobar or bilateral infiltrates on chest radiography.

I also want to comment on the relative elevation of her calcium, especially given the low albumin. This may simply be due to volume depletion, as many older patients have asymptomatic mild primary hyperparathyroidism. However, this elevated calcium may be a clue to the underlying lung process. Granulomatous lung disease due to tuberculosis or fungal infection could yield elevated calcium levels via increases in macrophage production of the active vitamin D metabolite calcitriol. This will need to be followed and a parathormone (PTH) level would be the best first test to request if the calcium level remains elevated. If the PTH level is suppressed, granulomatous disease or malignancy would be the more likely cause.

The patient was admitted with a presumptive diagnosis of community‐acquired pneumonia, was started on ceftriaxone and azithromycin, and given intravenous fluids, oxygen, and continued on inhaled salmeterol/fluticasone. Sputum was ordered for gram stain, culture, and sensitivity, and blood cultures were obtained. Urinalysis showed 1‐5 WBCs/high‐power field. Venous thromboembolism prophylaxis was initiated with subcutaneous heparin 5,000 units 8 hours. Her blood pressure normalized rapidly and during the next few days she stated she was feeling better. Despite continued significant wheezing her oxygen saturation remained at 98% on 2 L of oxygen via nasal cannula and she was less tachypneic. Attempts at obtaining an ABG were unsuccessful, and the patient subsequently refused additional attempts. Over the first few days her WBC count remained elevated above 20,000/L, with worsening bandemia (11%), and fever ranging from 38C to 39C. Sputum analysis was initially unsuccessful and blood cultures remained negative.

I am concerned about the persistent fever and elevated WBC count, and want to emphasize that I might have treated her with broader spectrum antibiotics to cover additional multidrug‐resistant bacterial organisms. I would have initially ordered vancomycin to cover methicillin resistant Staphylococcus aureus (MRSA) plus 2 additional antibiotics that cover multidrug‐resistant gram negative pathogens including Pseudomonas aeruginosa.

On the fifth hospital day, her WBC count dropped to 13,400/L and she defervesced. However, her respiratory status worsened during that same day with increased tachypnea. Of note, no results were reported from the initial sputum cultures and they were reordered and a noncontrast chest computed tomography (CT) was also ordered.

I think at this point, even though she has remained stable hemodynamically and oxygenating easily with supplemental oxygen, the question of whether or not her primary process is infectious or noninfectious lingers. I agree with obtaining a chest CT scan.

I am not surprised that sputum was not evaluated despite the orders. Among hospitalized patients with pneumonia, we frequently find that about a third of the time sputum cannot be obtained, about a third of the time it is obtained but the quality is unsatisfactory, and only a third of the time does the sputum sample meet criteria (less than 5 squamous epithelial cells per high‐power field) for adequate interpretation of the gram‐stain and culture result. Unfortunately, no one has developed a better way to improve this process. Nonetheless, I believe we do not try hard enough to obtain sputum in the first hours of evaluating our patients. I joke with our internal medicine residents that they should carry a sputum cup with them when they evaluate a patient with possible pneumonia. One recent prospective study of the value of sputum gram‐staining in community‐acquired pneumonia has found it to be highly specific for identifying Streptococcus pneumoniae or Haemophilus influenzae pneumonia.

The CT scan (Figure 2) performed on hospital day 6 demonstrated consolidation in the left upper lobe with areas of cavitation. There was also interstitial infiltrate extending into the lingula. Elevation of the right hemidiaphragm with atelectasis in both lung bases was also noted. A small effusion was present on the left and possibly a minimal effusion on the right as well. There was no pericardial effusion and only a few small pretracheal and periaortic lymph nodes were noted.

Figure 2

Given her failure to improve significantly after 6 days of antibiotic treatment, and her recent use of glucocorticoids, I would expand my diagnostic considerations to include other necrotizing bacterial infections, tuberculosis, fungus, and Nocardia.

Given the results of the CT scan she was placed in respiratory isolation to rule out active pulmonary tuberculosis. Though tachypneic, her blood pressure and pulse remained stable. However, her oxygen saturation deteriorated, declining to 92% on 2 L of oxygen via nasal cannula during hospital days 6 and 7. Subsequent successful attempts at collecting sputum yielded rapid growth of yeast (not Cryptococcus spp.). Pulmonary and infectious disease consultations were obtained and vancomycin was added to her regimen. The patient subsequently agreed to undergo diagnostic bronchoscopy.

I agree with obtaining input from expert consultants. I think we too often underutilize consultation in patients that are better but not completely better when we are not entirely sure what is going on. Evidence of noncryptococcus yeast in sputum may sometimes indicate colonization with Candida spp. without any significant clinical consequence. This finding may alternatively suggest the possibility of a true fungal pneumonia caused by 1 of the dimorphic fungi, including Histoplasma capsulatum, Paracoccidioides brasiliensis, Blastomyces dermatitides, or Coccidioides immitis. However, in this case there is not a strong epidemiologic patient history of exposure to any of these types of fungi.

Three sputum smears were negative for acid fast bacilli (AFB). Bronchoscopy revealed grossly abnormal mucosa in the left upper lobe and bronchomalacia, but no obstructive lesions. A transthoracic echocardiogram was ordered to evaluate her degree of pulmonary hypertension.

The 3 sputum specimens that were negative for AFB despite cavitary lung disease have high sensitivity for ruling out pulmonary tuberculosis. In addition, given the absence of any bacterial pathogen isolated from these specimens, I would pursue the possibility of other potential fungal pathogens given the patient's subacute course, history of using inhaled and oral corticosteroids, sputum results, and the presence of a cavitary lesion on her CT scan images.

Cytologic examination of the bronchoalveolar lavage (BAL) sample showed a cell differential of 1% bands, 58% neutrophils, 9% lymphs, and 27% eosinophils. The routine postbronchoscopy chest radiograph showed complete opacification of the left lung. The patient's WBC count rose to 26,000/L but she remained afebrile. Echocardiogram was reported to be of very poor quality due to her obesity. The cardiologist reviewing the echocardiogram called the attending physicians and stated there was possibly something in the left pulmonary artery and aortic dissection could not be ruled out.

The presence of eosinophilia on BAL may be a very important clue as to what lung pathology she has. In fact, eosinophilia in this setting may indicate the possibility of parasitic or fungal infection of the lung, or inflammation of the airway associated to drug toxicity, asthma, or environmental toxin exposure. With this additional information, I am concerned that she may be harboring an atypical infection such as an invasive fungus. The echocardiogram results are unclear to me but will need to be clarified with additional testing.

The interpretation of the transbronchial biopsy specimen was limited but suggested invasive pseudomembranous tracheal bronchitis due to aspergillosis. The routine hematoxylin and eosin stain showed portions of alveolar lung tissue and some collapsed submucosal bronchial glands with relatively normal‐looking lung tissue but along the edge of the spaces were obvious fungal organisms. The Gomori's methenamine silver (GMS) stain suggested the presence of Aspergillus organisms (Figure 3). Fungal cultures were also negative for any of the other dimorphic fungi or for molds.

Figure 3

Despite the negative culture results, the overall clinical picture suggests a necrotizing pneumonia caused by an invasive Aspergillus affecting both the bronchial tree and the lower respiratory tract. Generally, necrotizing pneumonias usually have a slow response to antimicrobial therapy. Given the inherent difficulty in differentiating clearly between invasive and noninvasive disease based on a transbronchial biopsy specimen, initiating antifungal therapy for invasive aspergillosis is appropriate in this patient. This patient's recent use of oral glucocorticoids and chronic use of inhaled glucocorticoids are both potential risk factors that predisposed this patient to develop invasive aspergillosis.

Many times we simply follow treatment guidelines for different categories of pneumonia, and have limited or inadequate clinical information to make more definitive diagnoses. While we need these treatment protocols, physicians must avoid falling into the trap that antibiotics treat all infectious etiologies in the lung and we should make reasonable efforts to pin down the etiology. All of us have been fooled by atypical presentations of tuberculosis, fungus, and noninfectious diseases of the lung. I think it behooves us to be vigilant about alternative diagnoses and consider pursuing additional studies whenever the clinical response to initial treatment does not meet our expectations.

Subsequently, the patient's additional cultures remained negative. The official echocardiogram report was read as questionable PE in the pulmonary artery. A spiral CT angiogram revealed a pulmonary artery embolus in the left upper lobe and she was treated with anticoagulation. Her shortness of breath improved steadily and she was successfully discharged after receiving 9 days of oral voriconazole. Outpatient pulmonary function testing documented the presence of chronic obstructive lung disease. She completed a 5‐month course of voriconazole therapy with significant clinical and radiologic improvement of her pulmonary infiltrate. She also completed a 12‐month treatment with warfarin for the concomitant pulmonary embolism. On follow‐up at 12 months she was doing well.

COMMENTARY

Aspergillosis caused particularly by Aspergillus fumigatus is considered an emerging infectious disease that frequently produces significant morbidity and mortality among immunocompromised patients.1, 2 The most frequently‐affected organs by this fungal pathogen include the lung and the central nervous system. There are 3 pathogenic mechanisms of Aspergillus infection of the lung: colonization, hypersensitivity reaction, and invasive aspergillosis.1

Invasive pulmonary aspergillosis is predominantly seen among individuals with severe degrees of immunosuppression as a result of solid‐organ transplantation, immunosuppressive therapies for autoimmune diseases, systemic glucocorticoids, and chemotherapy for hematologic malignancies. Mortality due to invasive aspergillosis continues to be very high (>58%) despite our improved ability to diagnose this condition and newer therapies to treat immunocompromised individuals.1 Invasive aspergillosis can manifest clinically in multiple ways. These include: (1) an invasive vascular process in which fungal organisms invade blood vessels, causing a rapidly progressive and often fatal illness; (2) necrotizing pseudomembranous tracheal bronchitis; (3) chronic necrotizing aspergillosis; (4) bronchopleural fistula; or (5) empyema.35 In our case, while the pathologic findings were most suggestive of an invasive pseudomembranous tracheal bronchitis, the overall clinical picture was most compatible with a necrotizing pneumonia due to invasive aspergillosis.

In addition to the traditional identified risk factors for invasive pulmonary aspergillosis, a number of reports during the last decade have demonstrated the occurrence of invasive aspergillosis in patients with COPD.14 A systematic review of the literature demonstrated that among 1,941 patients with invasive aspergillosis, 26 (1.3%) had evidence of COPD as the main risk factor for developing invasive aspergillosis.1 A single report has associated the potential use of inhaled steroids with the occurrence of invasive aspergillosis in this patient population.2 However, other factors that may promote increased susceptibility to invasive fungal infection among patients with COPD include the use of long‐term or repeated short‐term glucocorticoid treatments, and the presence of multiple additional comorbidities, which may be found in this same population such as diabetes, malnutrition, or end‐stage renal disease.3, 4 Most reported series have demonstrated a high mortality rate of invasive pulmonary aspergillosis in patients with COPD.14

The diagnosis of invasive pulmonary aspergillosis represents a significant clinical challenge. Diagnostic algorithms incorporating CT, antigen detection testing (for serum galactomannan and ‐glucan) as well as polymerase chain reaction diagnostic testing appear to be beneficial in the early diagnosis of invasive aspergillosis in particular settings such as in allogeneic hematopoietic stem cell transplantation.5 The role of antigen testing to identify early invasive aspergillosis in patients with COPD remains uncertain since it has been evaluated in a limited number of patients and therefore clinical suspicion is critical to push clinicians to pursue invasive tissue biopsy and cultures to confirm the diagnosis.3, 4

Based on the available clinical case series and in our case, invasive pulmonary aspergillosis should be suspected in COPD patients with rapidly progressive pneumonia not responding to antibacterial therapy and who have received oral or inhaled glucocorticoids in the recent past. In addition, this case also illustrates that occasionally, patients present with more than 1 life‐threatening diagnosis. This patient was also diagnosed with PE despite adequate prophylaxis. In addition to the well‐known clinical risk factors of obesity and lung disease, the underlying infection may have contributed to a systemic or local hypercoagulable condition that further increased her risk for venous thromboembolism.

KEY TEACHING POINTS

• Clinicians should remember to consider a broad differential in patients presenting with pneumonia, including the possibility of fungal pathogens in patients with known risk factors and in patients with multiple, potentially immunosuppressive comorbidities, or in patients who do not improve on standard antibiotic therapy.

• There is some evidence of an association between COPD and invasive aspergillosis, likely due to the frequent use of oral corticosteroids and/or chronic inhaled steroids in this population.

A 71‐year‐old African‐American woman presented to the emergency department with chest pain, shortness of breath, and cough. She had initially presented to her primary care physician 2 weeks previously complaining of worsening cough and shortness of breath and was told to continue her inhaled albuterol and glucocorticoids and was prescribed a prednisone taper and an unknown course of antibiotics. She noted no improvement in her symptoms despite compliance with this treatment. Three days prior to admission she described the gradual onset of left‐sided pleuritic chest pain with continued cough, associated with yellow sputum and worsening dyspnea. Review of systems was remarkable for generalized weakness and malaise. She denied fever, chills, orthopnea, paroxysmal nocturnal dyspnea, lower extremity edema, diarrhea, nausea, vomiting, or abdominal pain.

Her past medical history included a diagnosis of chronic obstructive pulmonary disease (COPD) but pulmonary function tests 7 years prior to admission showed an forced expiratory volume in the first second (FEV1)/forced vital capacity (FVC) ratio of 81%. She had a 30 pack‐year history of smoking, but quit 35 years ago. The patient also carried a diagnosis of heart failure, but an echocardiogram done 1 year ago demonstrated a left ventricular ejection fraction of 65% to 70% without diastolic dysfunction but mild right ventricular dilation and hypertrophy. Additionally, she had known nonobstructive coronary atherosclerotic heart disease, dyslipidemia, hypertension, morbid obesity, depression, and a documented chronic right hemidiaphragm elevation.

At this point the history suggests that the patient does not have a clear diagnosis of COPD. The lack of definitive spirometry evidence of chronic airway obstruction concerns me; I think that she may have been mistakenly treated with chronic inhaled steroids and doses of antibiotics for an acute exacerbation of chronic lung disease. Additional review of her history gives some indication of advanced lung disease, with her recent echocardiogram showing strain on the right ventricle with right ventricular hypertrophy and dilation, but there is no mention of the presence or severity of pulmonary hypertension. Nonetheless, I would be concerned that she probably has underlying significant cor pulmonale.

The patient now re‐presents with a worsening of her pulmonary symptoms. Her left‐sided pleuritic pain would make me concerned that she had a pulmonary embolus (PE). This morbidly obese patient with new pulmonary symptoms, right ventricular strain on her previous echocardiogram, and a persistent elevated right hemidiaphragm suggests a presentation of another PE.

At this time I cannot rule out other common possibilities such as infectious pneumonia. If she does have pneumonia, I would be concerned she could be harboring a multidrug‐resistant bacterial infection given her recent course of antibiotics in addition to her use of both chronic inhaled and intermittent oral glucocorticoids.

After gathering the rest of her full medical history, I would focus my physical exam on looking for evidence of parenchymal lung disease, signs of pulmonary hypertension, and pneumonia.

Her surgical history includes a previous hysterectomy, cholecystectomy, hernia repair, and left hepatic lobectomy for a benign mass. Her outpatient medications were ibuprofen, bupropion, fluvastatin, atenolol, potassium, aspirin, clopidogrel, albuterol inhaler, fluticasone/salmeterol inhaler, and omeprazole. She reports an allergy to penicillin and to sulfa drugs. Her mother died of an unknown cancer at age 77 years. She denied any international travel and she has always lived in Georgia.

The patient has been retired since 1992, having previously worked for the U.S. Postal Service. She admits to occasional alcohol intake (2 to 3 drinks a month). No recent travel, surgery, or prolonged immobilization was noted.

On initial examination she was alert and mentally appropriate, but appeared to be in mild respiratory distress with a respiratory rate of 28 breaths/minute. Her blood pressure (BP) was 99/70, heart rate 102, temperature of 38.2C, and oxygen saturation of 93% on room air and 97% on 2 L of oxygen via nasal cannula. Auscultation of her lungs revealed crackles over her left anterior lung field, bronchial breath sounds in the left posterior midlung, and bibasilar crackles. No wheezing was noted. Her cardiovascular exam and the remainder of her physical exam were unremarkable except for morbid obesity.

While my initial thoughts were leaning toward an exacerbation of chronic lung disease or possibly a new PE, at this moment, infection seems more likely. Indeed, her pulmonary findings suggest a left‐sided inflammatory process, and her vital signs meet criteria for systemic inflammatory response syndrome (SIRS). My primary concern is sepsis due to a drug‐resistant bacterial infection, including Staphylococcus aureus or gram‐negative bacteria or possibly more unusual organisms such as Nocardia or fungi, due to her recent use of antibiotics and chronic inhaled steroid use and recent course of oral glucocorticoids.

Conversely, the SIRS could be a manifestation of a noninfectious lung process such as acute interstitial pneumonia or an eosinophilic pneumonia. Given the diagnostic complexity, I would strongly consider consulting a pulmonologist if the patient did not improve quickly. At this point, I would like to review a posterior‐anterior (PA) and lateral chest radiograph, and room air arterial blood gas (ABG) in addition to basic laboratory test values.

Laboratory data obtained on admission was remarkable for a white blood cell (WBC) count of 26,500/L with 75% neutrophils and 6% eosinophils. Hemoglobin was 14.4 gm/dL. Platelet count was 454,000/L. Serum chemistries showed a sodium of 137 mEq/dL, potassium 4.3 mEq/dL, Cl 108 mEq/dL, bicarbonate 19 mEq/dL, blood urea nitrogen (BUN) 8 mg/dL, creatinine 1.0 mg/dL, and glucose 137 mg/dL. Cardiac enzymes were normal. Calcium was 9.8 mg/dL, albumin 2.7 gm/dL, total protein 6.9 gm/dL, AST 36 U/L, ALT 54 U/L and the bilirubin was normal. Chest radiograph (Figure 1) demonstrated a left perihilar infiltrate with air bronchograms and marked right hemidiaphragm elevation as seen on previous films. Unchanged increased interstitial markings were also present. Her electrocardiogram (ECG) showed normal sinus rhythm, normal axis, and QRS duration with nonspecific diffuse T‐wave abnormalities.

Figure 1

Given her presentation, I am worried about how well she is oxygenating and ventilating. An ABG should be done to assess her status more accurately. An albumin of 2.7 gm/dL indicates that she is fairly sick. I would not hesitate to consider testing the patient for human immunodeficiency virus (HIV) given how this information would dramatically change the differential diagnoses of her pulmonary process.

I am still most concerned about sepsis secondary to pneumonia in this patient with multiple chronic comorbidities, underlying chronic lung disease, receiving chronic inhaled glucocorticoids and a recent course of oral glucocorticoids and antibiotics. While I would initiate hydration I do not see a clear indication for early goal‐directed therapy for severe sepsis. In addition to obtaining an ABG and starting intravenous fluids, I would also draw blood cultures, send sputum for gram stain, culture, and sensitivity, and perform a urinalysis. I would also administer empiric antibiotics as quickly as possible based on a number of pneumonia clinical studies suggesting improved outcomes with early antibiotic administration. Because of her use of antibiotics and both inhaled and oral glucocorticoids, she is at higher risk for potentially multidrug‐resistant bacterial pathogens, including Staphyloccocus aureus and gram‐negative bacteria such as Pseudomonas and Klebsiella (Table 1). Therefore, I would initially cover her broadly for these organisms.

In addition to initial treatment choice, the inpatient triage decision is another important issue, especially at a community hospital where intensive care unit (ICU) resources are rare and often the admission decision is between sending a moderately sick patient to a regular floor bed or the medical ICU. Both the American Thoracic Society and Infectious Diseases Society of America support an ICU triage protocol in their guidelines for the management of community‐acquired pneumonia in adults that utilizes the following 9 minor criteria, of which the presence of at least 3 should support ICU admission: respiratory rate 30 breaths/minute; oxygenation index (pressure of oxygen [PaO₂]/fraction of inspired oxygen [FiO₂] ratio) 250; multilobar infiltrates; confusion/disorientation; uremia (BUN level 20 mg/dL); leukopenia (WBC count <4,000 cells/mm³); thrombocytopenia (platelet count <100,000 cells/mm³); hypothermia (core temperature <36C); and hypotension requiring aggressive fluid. Despite the absence of these criteria in this patient, it is important to note that no triage protocol has been adequately prospectively validated. Retrospective study of the minor criteria has found that the presence of at least 2 of the following 3 clinical criteria to have the highest specificity for predicting cardiopulmonary decompensation and subsequent need for ICU care: (1) initial hypotension (BP <90/60) on presentation with response to initial intravenous fluids to a BP >90/60; (2) oxygenation failure as indicated by PaO₂/FiO₂ ratio less than 250; or (3) the presence of multilobar or bilateral infiltrates on chest radiography.

I also want to comment on the relative elevation of her calcium, especially given the low albumin. This may simply be due to volume depletion, as many older patients have asymptomatic mild primary hyperparathyroidism. However, this elevated calcium may be a clue to the underlying lung process. Granulomatous lung disease due to tuberculosis or fungal infection could yield elevated calcium levels via increases in macrophage production of the active vitamin D metabolite calcitriol. This will need to be followed and a parathormone (PTH) level would be the best first test to request if the calcium level remains elevated. If the PTH level is suppressed, granulomatous disease or malignancy would be the more likely cause.

The patient was admitted with a presumptive diagnosis of community‐acquired pneumonia, was started on ceftriaxone and azithromycin, and given intravenous fluids, oxygen, and continued on inhaled salmeterol/fluticasone. Sputum was ordered for gram stain, culture, and sensitivity, and blood cultures were obtained. Urinalysis showed 1‐5 WBCs/high‐power field. Venous thromboembolism prophylaxis was initiated with subcutaneous heparin 5,000 units 8 hours. Her blood pressure normalized rapidly and during the next few days she stated she was feeling better. Despite continued significant wheezing her oxygen saturation remained at 98% on 2 L of oxygen via nasal cannula and she was less tachypneic. Attempts at obtaining an ABG were unsuccessful, and the patient subsequently refused additional attempts. Over the first few days her WBC count remained elevated above 20,000/L, with worsening bandemia (11%), and fever ranging from 38C to 39C. Sputum analysis was initially unsuccessful and blood cultures remained negative.

I am concerned about the persistent fever and elevated WBC count, and want to emphasize that I might have treated her with broader spectrum antibiotics to cover additional multidrug‐resistant bacterial organisms. I would have initially ordered vancomycin to cover methicillin resistant Staphylococcus aureus (MRSA) plus 2 additional antibiotics that cover multidrug‐resistant gram negative pathogens including Pseudomonas aeruginosa.

On the fifth hospital day, her WBC count dropped to 13,400/L and she defervesced. However, her respiratory status worsened during that same day with increased tachypnea. Of note, no results were reported from the initial sputum cultures and they were reordered and a noncontrast chest computed tomography (CT) was also ordered.

I think at this point, even though she has remained stable hemodynamically and oxygenating easily with supplemental oxygen, the question of whether or not her primary process is infectious or noninfectious lingers. I agree with obtaining a chest CT scan.

I am not surprised that sputum was not evaluated despite the orders. Among hospitalized patients with pneumonia, we frequently find that about a third of the time sputum cannot be obtained, about a third of the time it is obtained but the quality is unsatisfactory, and only a third of the time does the sputum sample meet criteria (less than 5 squamous epithelial cells per high‐power field) for adequate interpretation of the gram‐stain and culture result. Unfortunately, no one has developed a better way to improve this process. Nonetheless, I believe we do not try hard enough to obtain sputum in the first hours of evaluating our patients. I joke with our internal medicine residents that they should carry a sputum cup with them when they evaluate a patient with possible pneumonia. One recent prospective study of the value of sputum gram‐staining in community‐acquired pneumonia has found it to be highly specific for identifying Streptococcus pneumoniae or Haemophilus influenzae pneumonia.

The CT scan (Figure 2) performed on hospital day 6 demonstrated consolidation in the left upper lobe with areas of cavitation. There was also interstitial infiltrate extending into the lingula. Elevation of the right hemidiaphragm with atelectasis in both lung bases was also noted. A small effusion was present on the left and possibly a minimal effusion on the right as well. There was no pericardial effusion and only a few small pretracheal and periaortic lymph nodes were noted.

Figure 2

Given her failure to improve significantly after 6 days of antibiotic treatment, and her recent use of glucocorticoids, I would expand my diagnostic considerations to include other necrotizing bacterial infections, tuberculosis, fungus, and Nocardia.

Given the results of the CT scan she was placed in respiratory isolation to rule out active pulmonary tuberculosis. Though tachypneic, her blood pressure and pulse remained stable. However, her oxygen saturation deteriorated, declining to 92% on 2 L of oxygen via nasal cannula during hospital days 6 and 7. Subsequent successful attempts at collecting sputum yielded rapid growth of yeast (not Cryptococcus spp.). Pulmonary and infectious disease consultations were obtained and vancomycin was added to her regimen. The patient subsequently agreed to undergo diagnostic bronchoscopy.

I agree with obtaining input from expert consultants. I think we too often underutilize consultation in patients that are better but not completely better when we are not entirely sure what is going on. Evidence of noncryptococcus yeast in sputum may sometimes indicate colonization with Candida spp. without any significant clinical consequence. This finding may alternatively suggest the possibility of a true fungal pneumonia caused by 1 of the dimorphic fungi, including Histoplasma capsulatum, Paracoccidioides brasiliensis, Blastomyces dermatitides, or Coccidioides immitis. However, in this case there is not a strong epidemiologic patient history of exposure to any of these types of fungi.

Three sputum smears were negative for acid fast bacilli (AFB). Bronchoscopy revealed grossly abnormal mucosa in the left upper lobe and bronchomalacia, but no obstructive lesions. A transthoracic echocardiogram was ordered to evaluate her degree of pulmonary hypertension.

The 3 sputum specimens that were negative for AFB despite cavitary lung disease have high sensitivity for ruling out pulmonary tuberculosis. In addition, given the absence of any bacterial pathogen isolated from these specimens, I would pursue the possibility of other potential fungal pathogens given the patient's subacute course, history of using inhaled and oral corticosteroids, sputum results, and the presence of a cavitary lesion on her CT scan images.

Cytologic examination of the bronchoalveolar lavage (BAL) sample showed a cell differential of 1% bands, 58% neutrophils, 9% lymphs, and 27% eosinophils. The routine postbronchoscopy chest radiograph showed complete opacification of the left lung. The patient's WBC count rose to 26,000/L but she remained afebrile. Echocardiogram was reported to be of very poor quality due to her obesity. The cardiologist reviewing the echocardiogram called the attending physicians and stated there was possibly something in the left pulmonary artery and aortic dissection could not be ruled out.

The presence of eosinophilia on BAL may be a very important clue as to what lung pathology she has. In fact, eosinophilia in this setting may indicate the possibility of parasitic or fungal infection of the lung, or inflammation of the airway associated to drug toxicity, asthma, or environmental toxin exposure. With this additional information, I am concerned that she may be harboring an atypical infection such as an invasive fungus. The echocardiogram results are unclear to me but will need to be clarified with additional testing.

The interpretation of the transbronchial biopsy specimen was limited but suggested invasive pseudomembranous tracheal bronchitis due to aspergillosis. The routine hematoxylin and eosin stain showed portions of alveolar lung tissue and some collapsed submucosal bronchial glands with relatively normal‐looking lung tissue but along the edge of the spaces were obvious fungal organisms. The Gomori's methenamine silver (GMS) stain suggested the presence of Aspergillus organisms (Figure 3). Fungal cultures were also negative for any of the other dimorphic fungi or for molds.

Figure 3

Despite the negative culture results, the overall clinical picture suggests a necrotizing pneumonia caused by an invasive Aspergillus affecting both the bronchial tree and the lower respiratory tract. Generally, necrotizing pneumonias usually have a slow response to antimicrobial therapy. Given the inherent difficulty in differentiating clearly between invasive and noninvasive disease based on a transbronchial biopsy specimen, initiating antifungal therapy for invasive aspergillosis is appropriate in this patient. This patient's recent use of oral glucocorticoids and chronic use of inhaled glucocorticoids are both potential risk factors that predisposed this patient to develop invasive aspergillosis.

Many times we simply follow treatment guidelines for different categories of pneumonia, and have limited or inadequate clinical information to make more definitive diagnoses. While we need these treatment protocols, physicians must avoid falling into the trap that antibiotics treat all infectious etiologies in the lung and we should make reasonable efforts to pin down the etiology. All of us have been fooled by atypical presentations of tuberculosis, fungus, and noninfectious diseases of the lung. I think it behooves us to be vigilant about alternative diagnoses and consider pursuing additional studies whenever the clinical response to initial treatment does not meet our expectations.

Subsequently, the patient's additional cultures remained negative. The official echocardiogram report was read as questionable PE in the pulmonary artery. A spiral CT angiogram revealed a pulmonary artery embolus in the left upper lobe and she was treated with anticoagulation. Her shortness of breath improved steadily and she was successfully discharged after receiving 9 days of oral voriconazole. Outpatient pulmonary function testing documented the presence of chronic obstructive lung disease. She completed a 5‐month course of voriconazole therapy with significant clinical and radiologic improvement of her pulmonary infiltrate. She also completed a 12‐month treatment with warfarin for the concomitant pulmonary embolism. On follow‐up at 12 months she was doing well.

COMMENTARY

Aspergillosis caused particularly by Aspergillus fumigatus is considered an emerging infectious disease that frequently produces significant morbidity and mortality among immunocompromised patients.1, 2 The most frequently‐affected organs by this fungal pathogen include the lung and the central nervous system. There are 3 pathogenic mechanisms of Aspergillus infection of the lung: colonization, hypersensitivity reaction, and invasive aspergillosis.1

Invasive pulmonary aspergillosis is predominantly seen among individuals with severe degrees of immunosuppression as a result of solid‐organ transplantation, immunosuppressive therapies for autoimmune diseases, systemic glucocorticoids, and chemotherapy for hematologic malignancies. Mortality due to invasive aspergillosis continues to be very high (>58%) despite our improved ability to diagnose this condition and newer therapies to treat immunocompromised individuals.1 Invasive aspergillosis can manifest clinically in multiple ways. These include: (1) an invasive vascular process in which fungal organisms invade blood vessels, causing a rapidly progressive and often fatal illness; (2) necrotizing pseudomembranous tracheal bronchitis; (3) chronic necrotizing aspergillosis; (4) bronchopleural fistula; or (5) empyema.35 In our case, while the pathologic findings were most suggestive of an invasive pseudomembranous tracheal bronchitis, the overall clinical picture was most compatible with a necrotizing pneumonia due to invasive aspergillosis.

In addition to the traditional identified risk factors for invasive pulmonary aspergillosis, a number of reports during the last decade have demonstrated the occurrence of invasive aspergillosis in patients with COPD.14 A systematic review of the literature demonstrated that among 1,941 patients with invasive aspergillosis, 26 (1.3%) had evidence of COPD as the main risk factor for developing invasive aspergillosis.1 A single report has associated the potential use of inhaled steroids with the occurrence of invasive aspergillosis in this patient population.2 However, other factors that may promote increased susceptibility to invasive fungal infection among patients with COPD include the use of long‐term or repeated short‐term glucocorticoid treatments, and the presence of multiple additional comorbidities, which may be found in this same population such as diabetes, malnutrition, or end‐stage renal disease.3, 4 Most reported series have demonstrated a high mortality rate of invasive pulmonary aspergillosis in patients with COPD.14

The diagnosis of invasive pulmonary aspergillosis represents a significant clinical challenge. Diagnostic algorithms incorporating CT, antigen detection testing (for serum galactomannan and ‐glucan) as well as polymerase chain reaction diagnostic testing appear to be beneficial in the early diagnosis of invasive aspergillosis in particular settings such as in allogeneic hematopoietic stem cell transplantation.5 The role of antigen testing to identify early invasive aspergillosis in patients with COPD remains uncertain since it has been evaluated in a limited number of patients and therefore clinical suspicion is critical to push clinicians to pursue invasive tissue biopsy and cultures to confirm the diagnosis.3, 4

Based on the available clinical case series and in our case, invasive pulmonary aspergillosis should be suspected in COPD patients with rapidly progressive pneumonia not responding to antibacterial therapy and who have received oral or inhaled glucocorticoids in the recent past. In addition, this case also illustrates that occasionally, patients present with more than 1 life‐threatening diagnosis. This patient was also diagnosed with PE despite adequate prophylaxis. In addition to the well‐known clinical risk factors of obesity and lung disease, the underlying infection may have contributed to a systemic or local hypercoagulable condition that further increased her risk for venous thromboembolism.

KEY TEACHING POINTS

• Clinicians should remember to consider a broad differential in patients presenting with pneumonia, including the possibility of fungal pathogens in patients with known risk factors and in patients with multiple, potentially immunosuppressive comorbidities, or in patients who do not improve on standard antibiotic therapy.

• There is some evidence of an association between COPD and invasive aspergillosis, likely due to the frequent use of oral corticosteroids and/or chronic inhaled steroids in this population.

A 71‐year‐old African‐American woman presented to the emergency department with chest pain, shortness of breath, and cough. She had initially presented to her primary care physician 2 weeks previously complaining of worsening cough and shortness of breath and was told to continue her inhaled albuterol and glucocorticoids and was prescribed a prednisone taper and an unknown course of antibiotics. She noted no improvement in her symptoms despite compliance with this treatment. Three days prior to admission she described the gradual onset of left‐sided pleuritic chest pain with continued cough, associated with yellow sputum and worsening dyspnea. Review of systems was remarkable for generalized weakness and malaise. She denied fever, chills, orthopnea, paroxysmal nocturnal dyspnea, lower extremity edema, diarrhea, nausea, vomiting, or abdominal pain.

Her past medical history included a diagnosis of chronic obstructive pulmonary disease (COPD) but pulmonary function tests 7 years prior to admission showed an forced expiratory volume in the first second (FEV1)/forced vital capacity (FVC) ratio of 81%. She had a 30 pack‐year history of smoking, but quit 35 years ago. The patient also carried a diagnosis of heart failure, but an echocardiogram done 1 year ago demonstrated a left ventricular ejection fraction of 65% to 70% without diastolic dysfunction but mild right ventricular dilation and hypertrophy. Additionally, she had known nonobstructive coronary atherosclerotic heart disease, dyslipidemia, hypertension, morbid obesity, depression, and a documented chronic right hemidiaphragm elevation.

At this point the history suggests that the patient does not have a clear diagnosis of COPD. The lack of definitive spirometry evidence of chronic airway obstruction concerns me; I think that she may have been mistakenly treated with chronic inhaled steroids and doses of antibiotics for an acute exacerbation of chronic lung disease. Additional review of her history gives some indication of advanced lung disease, with her recent echocardiogram showing strain on the right ventricle with right ventricular hypertrophy and dilation, but there is no mention of the presence or severity of pulmonary hypertension. Nonetheless, I would be concerned that she probably has underlying significant cor pulmonale.

The patient now re‐presents with a worsening of her pulmonary symptoms. Her left‐sided pleuritic pain would make me concerned that she had a pulmonary embolus (PE). This morbidly obese patient with new pulmonary symptoms, right ventricular strain on her previous echocardiogram, and a persistent elevated right hemidiaphragm suggests a presentation of another PE.

At this time I cannot rule out other common possibilities such as infectious pneumonia. If she does have pneumonia, I would be concerned she could be harboring a multidrug‐resistant bacterial infection given her recent course of antibiotics in addition to her use of both chronic inhaled and intermittent oral glucocorticoids.

After gathering the rest of her full medical history, I would focus my physical exam on looking for evidence of parenchymal lung disease, signs of pulmonary hypertension, and pneumonia.

Her surgical history includes a previous hysterectomy, cholecystectomy, hernia repair, and left hepatic lobectomy for a benign mass. Her outpatient medications were ibuprofen, bupropion, fluvastatin, atenolol, potassium, aspirin, clopidogrel, albuterol inhaler, fluticasone/salmeterol inhaler, and omeprazole. She reports an allergy to penicillin and to sulfa drugs. Her mother died of an unknown cancer at age 77 years. She denied any international travel and she has always lived in Georgia.

The patient has been retired since 1992, having previously worked for the U.S. Postal Service. She admits to occasional alcohol intake (2 to 3 drinks a month). No recent travel, surgery, or prolonged immobilization was noted.

On initial examination she was alert and mentally appropriate, but appeared to be in mild respiratory distress with a respiratory rate of 28 breaths/minute. Her blood pressure (BP) was 99/70, heart rate 102, temperature of 38.2C, and oxygen saturation of 93% on room air and 97% on 2 L of oxygen via nasal cannula. Auscultation of her lungs revealed crackles over her left anterior lung field, bronchial breath sounds in the left posterior midlung, and bibasilar crackles. No wheezing was noted. Her cardiovascular exam and the remainder of her physical exam were unremarkable except for morbid obesity.

While my initial thoughts were leaning toward an exacerbation of chronic lung disease or possibly a new PE, at this moment, infection seems more likely. Indeed, her pulmonary findings suggest a left‐sided inflammatory process, and her vital signs meet criteria for systemic inflammatory response syndrome (SIRS). My primary concern is sepsis due to a drug‐resistant bacterial infection, including Staphylococcus aureus or gram‐negative bacteria or possibly more unusual organisms such as Nocardia or fungi, due to her recent use of antibiotics and chronic inhaled steroid use and recent course of oral glucocorticoids.

Conversely, the SIRS could be a manifestation of a noninfectious lung process such as acute interstitial pneumonia or an eosinophilic pneumonia. Given the diagnostic complexity, I would strongly consider consulting a pulmonologist if the patient did not improve quickly. At this point, I would like to review a posterior‐anterior (PA) and lateral chest radiograph, and room air arterial blood gas (ABG) in addition to basic laboratory test values.

Laboratory data obtained on admission was remarkable for a white blood cell (WBC) count of 26,500/L with 75% neutrophils and 6% eosinophils. Hemoglobin was 14.4 gm/dL. Platelet count was 454,000/L. Serum chemistries showed a sodium of 137 mEq/dL, potassium 4.3 mEq/dL, Cl 108 mEq/dL, bicarbonate 19 mEq/dL, blood urea nitrogen (BUN) 8 mg/dL, creatinine 1.0 mg/dL, and glucose 137 mg/dL. Cardiac enzymes were normal. Calcium was 9.8 mg/dL, albumin 2.7 gm/dL, total protein 6.9 gm/dL, AST 36 U/L, ALT 54 U/L and the bilirubin was normal. Chest radiograph (Figure 1) demonstrated a left perihilar infiltrate with air bronchograms and marked right hemidiaphragm elevation as seen on previous films. Unchanged increased interstitial markings were also present. Her electrocardiogram (ECG) showed normal sinus rhythm, normal axis, and QRS duration with nonspecific diffuse T‐wave abnormalities.

Figure 1

Given her presentation, I am worried about how well she is oxygenating and ventilating. An ABG should be done to assess her status more accurately. An albumin of 2.7 gm/dL indicates that she is fairly sick. I would not hesitate to consider testing the patient for human immunodeficiency virus (HIV) given how this information would dramatically change the differential diagnoses of her pulmonary process.

I am still most concerned about sepsis secondary to pneumonia in this patient with multiple chronic comorbidities, underlying chronic lung disease, receiving chronic inhaled glucocorticoids and a recent course of oral glucocorticoids and antibiotics. While I would initiate hydration I do not see a clear indication for early goal‐directed therapy for severe sepsis. In addition to obtaining an ABG and starting intravenous fluids, I would also draw blood cultures, send sputum for gram stain, culture, and sensitivity, and perform a urinalysis. I would also administer empiric antibiotics as quickly as possible based on a number of pneumonia clinical studies suggesting improved outcomes with early antibiotic administration. Because of her use of antibiotics and both inhaled and oral glucocorticoids, she is at higher risk for potentially multidrug‐resistant bacterial pathogens, including Staphyloccocus aureus and gram‐negative bacteria such as Pseudomonas and Klebsiella (Table 1). Therefore, I would initially cover her broadly for these organisms.

In addition to initial treatment choice, the inpatient triage decision is another important issue, especially at a community hospital where intensive care unit (ICU) resources are rare and often the admission decision is between sending a moderately sick patient to a regular floor bed or the medical ICU. Both the American Thoracic Society and Infectious Diseases Society of America support an ICU triage protocol in their guidelines for the management of community‐acquired pneumonia in adults that utilizes the following 9 minor criteria, of which the presence of at least 3 should support ICU admission: respiratory rate 30 breaths/minute; oxygenation index (pressure of oxygen [PaO₂]/fraction of inspired oxygen [FiO₂] ratio) 250; multilobar infiltrates; confusion/disorientation; uremia (BUN level 20 mg/dL); leukopenia (WBC count <4,000 cells/mm³); thrombocytopenia (platelet count <100,000 cells/mm³); hypothermia (core temperature <36C); and hypotension requiring aggressive fluid. Despite the absence of these criteria in this patient, it is important to note that no triage protocol has been adequately prospectively validated. Retrospective study of the minor criteria has found that the presence of at least 2 of the following 3 clinical criteria to have the highest specificity for predicting cardiopulmonary decompensation and subsequent need for ICU care: (1) initial hypotension (BP <90/60) on presentation with response to initial intravenous fluids to a BP >90/60; (2) oxygenation failure as indicated by PaO₂/FiO₂ ratio less than 250; or (3) the presence of multilobar or bilateral infiltrates on chest radiography.

I also want to comment on the relative elevation of her calcium, especially given the low albumin. This may simply be due to volume depletion, as many older patients have asymptomatic mild primary hyperparathyroidism. However, this elevated calcium may be a clue to the underlying lung process. Granulomatous lung disease due to tuberculosis or fungal infection could yield elevated calcium levels via increases in macrophage production of the active vitamin D metabolite calcitriol. This will need to be followed and a parathormone (PTH) level would be the best first test to request if the calcium level remains elevated. If the PTH level is suppressed, granulomatous disease or malignancy would be the more likely cause.

The patient was admitted with a presumptive diagnosis of community‐acquired pneumonia, was started on ceftriaxone and azithromycin, and given intravenous fluids, oxygen, and continued on inhaled salmeterol/fluticasone. Sputum was ordered for gram stain, culture, and sensitivity, and blood cultures were obtained. Urinalysis showed 1‐5 WBCs/high‐power field. Venous thromboembolism prophylaxis was initiated with subcutaneous heparin 5,000 units 8 hours. Her blood pressure normalized rapidly and during the next few days she stated she was feeling better. Despite continued significant wheezing her oxygen saturation remained at 98% on 2 L of oxygen via nasal cannula and she was less tachypneic. Attempts at obtaining an ABG were unsuccessful, and the patient subsequently refused additional attempts. Over the first few days her WBC count remained elevated above 20,000/L, with worsening bandemia (11%), and fever ranging from 38C to 39C. Sputum analysis was initially unsuccessful and blood cultures remained negative.

I am concerned about the persistent fever and elevated WBC count, and want to emphasize that I might have treated her with broader spectrum antibiotics to cover additional multidrug‐resistant bacterial organisms. I would have initially ordered vancomycin to cover methicillin resistant Staphylococcus aureus (MRSA) plus 2 additional antibiotics that cover multidrug‐resistant gram negative pathogens including Pseudomonas aeruginosa.

On the fifth hospital day, her WBC count dropped to 13,400/L and she defervesced. However, her respiratory status worsened during that same day with increased tachypnea. Of note, no results were reported from the initial sputum cultures and they were reordered and a noncontrast chest computed tomography (CT) was also ordered.

I think at this point, even though she has remained stable hemodynamically and oxygenating easily with supplemental oxygen, the question of whether or not her primary process is infectious or noninfectious lingers. I agree with obtaining a chest CT scan.

I am not surprised that sputum was not evaluated despite the orders. Among hospitalized patients with pneumonia, we frequently find that about a third of the time sputum cannot be obtained, about a third of the time it is obtained but the quality is unsatisfactory, and only a third of the time does the sputum sample meet criteria (less than 5 squamous epithelial cells per high‐power field) for adequate interpretation of the gram‐stain and culture result. Unfortunately, no one has developed a better way to improve this process. Nonetheless, I believe we do not try hard enough to obtain sputum in the first hours of evaluating our patients. I joke with our internal medicine residents that they should carry a sputum cup with them when they evaluate a patient with possible pneumonia. One recent prospective study of the value of sputum gram‐staining in community‐acquired pneumonia has found it to be highly specific for identifying Streptococcus pneumoniae or Haemophilus influenzae pneumonia.

The CT scan (Figure 2) performed on hospital day 6 demonstrated consolidation in the left upper lobe with areas of cavitation. There was also interstitial infiltrate extending into the lingula. Elevation of the right hemidiaphragm with atelectasis in both lung bases was also noted. A small effusion was present on the left and possibly a minimal effusion on the right as well. There was no pericardial effusion and only a few small pretracheal and periaortic lymph nodes were noted.

Figure 2

Given her failure to improve significantly after 6 days of antibiotic treatment, and her recent use of glucocorticoids, I would expand my diagnostic considerations to include other necrotizing bacterial infections, tuberculosis, fungus, and Nocardia.

Given the results of the CT scan she was placed in respiratory isolation to rule out active pulmonary tuberculosis. Though tachypneic, her blood pressure and pulse remained stable. However, her oxygen saturation deteriorated, declining to 92% on 2 L of oxygen via nasal cannula during hospital days 6 and 7. Subsequent successful attempts at collecting sputum yielded rapid growth of yeast (not Cryptococcus spp.). Pulmonary and infectious disease consultations were obtained and vancomycin was added to her regimen. The patient subsequently agreed to undergo diagnostic bronchoscopy.

I agree with obtaining input from expert consultants. I think we too often underutilize consultation in patients that are better but not completely better when we are not entirely sure what is going on. Evidence of noncryptococcus yeast in sputum may sometimes indicate colonization with Candida spp. without any significant clinical consequence. This finding may alternatively suggest the possibility of a true fungal pneumonia caused by 1 of the dimorphic fungi, including Histoplasma capsulatum, Paracoccidioides brasiliensis, Blastomyces dermatitides, or Coccidioides immitis. However, in this case there is not a strong epidemiologic patient history of exposure to any of these types of fungi.

Three sputum smears were negative for acid fast bacilli (AFB). Bronchoscopy revealed grossly abnormal mucosa in the left upper lobe and bronchomalacia, but no obstructive lesions. A transthoracic echocardiogram was ordered to evaluate her degree of pulmonary hypertension.

The 3 sputum specimens that were negative for AFB despite cavitary lung disease have high sensitivity for ruling out pulmonary tuberculosis. In addition, given the absence of any bacterial pathogen isolated from these specimens, I would pursue the possibility of other potential fungal pathogens given the patient's subacute course, history of using inhaled and oral corticosteroids, sputum results, and the presence of a cavitary lesion on her CT scan images.

Cytologic examination of the bronchoalveolar lavage (BAL) sample showed a cell differential of 1% bands, 58% neutrophils, 9% lymphs, and 27% eosinophils. The routine postbronchoscopy chest radiograph showed complete opacification of the left lung. The patient's WBC count rose to 26,000/L but she remained afebrile. Echocardiogram was reported to be of very poor quality due to her obesity. The cardiologist reviewing the echocardiogram called the attending physicians and stated there was possibly something in the left pulmonary artery and aortic dissection could not be ruled out.

The presence of eosinophilia on BAL may be a very important clue as to what lung pathology she has. In fact, eosinophilia in this setting may indicate the possibility of parasitic or fungal infection of the lung, or inflammation of the airway associated to drug toxicity, asthma, or environmental toxin exposure. With this additional information, I am concerned that she may be harboring an atypical infection such as an invasive fungus. The echocardiogram results are unclear to me but will need to be clarified with additional testing.

The interpretation of the transbronchial biopsy specimen was limited but suggested invasive pseudomembranous tracheal bronchitis due to aspergillosis. The routine hematoxylin and eosin stain showed portions of alveolar lung tissue and some collapsed submucosal bronchial glands with relatively normal‐looking lung tissue but along the edge of the spaces were obvious fungal organisms. The Gomori's methenamine silver (GMS) stain suggested the presence of Aspergillus organisms (Figure 3). Fungal cultures were also negative for any of the other dimorphic fungi or for molds.

Figure 3

Despite the negative culture results, the overall clinical picture suggests a necrotizing pneumonia caused by an invasive Aspergillus affecting both the bronchial tree and the lower respiratory tract. Generally, necrotizing pneumonias usually have a slow response to antimicrobial therapy. Given the inherent difficulty in differentiating clearly between invasive and noninvasive disease based on a transbronchial biopsy specimen, initiating antifungal therapy for invasive aspergillosis is appropriate in this patient. This patient's recent use of oral glucocorticoids and chronic use of inhaled glucocorticoids are both potential risk factors that predisposed this patient to develop invasive aspergillosis.

Many times we simply follow treatment guidelines for different categories of pneumonia, and have limited or inadequate clinical information to make more definitive diagnoses. While we need these treatment protocols, physicians must avoid falling into the trap that antibiotics treat all infectious etiologies in the lung and we should make reasonable efforts to pin down the etiology. All of us have been fooled by atypical presentations of tuberculosis, fungus, and noninfectious diseases of the lung. I think it behooves us to be vigilant about alternative diagnoses and consider pursuing additional studies whenever the clinical response to initial treatment does not meet our expectations.

Subsequently, the patient's additional cultures remained negative. The official echocardiogram report was read as questionable PE in the pulmonary artery. A spiral CT angiogram revealed a pulmonary artery embolus in the left upper lobe and she was treated with anticoagulation. Her shortness of breath improved steadily and she was successfully discharged after receiving 9 days of oral voriconazole. Outpatient pulmonary function testing documented the presence of chronic obstructive lung disease. She completed a 5‐month course of voriconazole therapy with significant clinical and radiologic improvement of her pulmonary infiltrate. She also completed a 12‐month treatment with warfarin for the concomitant pulmonary embolism. On follow‐up at 12 months she was doing well.

COMMENTARY

Aspergillosis caused particularly by Aspergillus fumigatus is considered an emerging infectious disease that frequently produces significant morbidity and mortality among immunocompromised patients.1, 2 The most frequently‐affected organs by this fungal pathogen include the lung and the central nervous system. There are 3 pathogenic mechanisms of Aspergillus infection of the lung: colonization, hypersensitivity reaction, and invasive aspergillosis.1

Invasive pulmonary aspergillosis is predominantly seen among individuals with severe degrees of immunosuppression as a result of solid‐organ transplantation, immunosuppressive therapies for autoimmune diseases, systemic glucocorticoids, and chemotherapy for hematologic malignancies. Mortality due to invasive aspergillosis continues to be very high (>58%) despite our improved ability to diagnose this condition and newer therapies to treat immunocompromised individuals.1 Invasive aspergillosis can manifest clinically in multiple ways. These include: (1) an invasive vascular process in which fungal organisms invade blood vessels, causing a rapidly progressive and often fatal illness; (2) necrotizing pseudomembranous tracheal bronchitis; (3) chronic necrotizing aspergillosis; (4) bronchopleural fistula; or (5) empyema.35 In our case, while the pathologic findings were most suggestive of an invasive pseudomembranous tracheal bronchitis, the overall clinical picture was most compatible with a necrotizing pneumonia due to invasive aspergillosis.

In addition to the traditional identified risk factors for invasive pulmonary aspergillosis, a number of reports during the last decade have demonstrated the occurrence of invasive aspergillosis in patients with COPD.14 A systematic review of the literature demonstrated that among 1,941 patients with invasive aspergillosis, 26 (1.3%) had evidence of COPD as the main risk factor for developing invasive aspergillosis.1 A single report has associated the potential use of inhaled steroids with the occurrence of invasive aspergillosis in this patient population.2 However, other factors that may promote increased susceptibility to invasive fungal infection among patients with COPD include the use of long‐term or repeated short‐term glucocorticoid treatments, and the presence of multiple additional comorbidities, which may be found in this same population such as diabetes, malnutrition, or end‐stage renal disease.3, 4 Most reported series have demonstrated a high mortality rate of invasive pulmonary aspergillosis in patients with COPD.14

The diagnosis of invasive pulmonary aspergillosis represents a significant clinical challenge. Diagnostic algorithms incorporating CT, antigen detection testing (for serum galactomannan and ‐glucan) as well as polymerase chain reaction diagnostic testing appear to be beneficial in the early diagnosis of invasive aspergillosis in particular settings such as in allogeneic hematopoietic stem cell transplantation.5 The role of antigen testing to identify early invasive aspergillosis in patients with COPD remains uncertain since it has been evaluated in a limited number of patients and therefore clinical suspicion is critical to push clinicians to pursue invasive tissue biopsy and cultures to confirm the diagnosis.3, 4

Based on the available clinical case series and in our case, invasive pulmonary aspergillosis should be suspected in COPD patients with rapidly progressive pneumonia not responding to antibacterial therapy and who have received oral or inhaled glucocorticoids in the recent past. In addition, this case also illustrates that occasionally, patients present with more than 1 life‐threatening diagnosis. This patient was also diagnosed with PE despite adequate prophylaxis. In addition to the well‐known clinical risk factors of obesity and lung disease, the underlying infection may have contributed to a systemic or local hypercoagulable condition that further increased her risk for venous thromboembolism.

KEY TEACHING POINTS

• Clinicians should remember to consider a broad differential in patients presenting with pneumonia, including the possibility of fungal pathogens in patients with known risk factors and in patients with multiple, potentially immunosuppressive comorbidities, or in patients who do not improve on standard antibiotic therapy.

• There is some evidence of an association between COPD and invasive aspergillosis, likely due to the frequent use of oral corticosteroids and/or chronic inhaled steroids in this population.

The Joint Commission's (TJC) National Patient Safety Goal (NPSG) #8Accurately and completely reconcile medications across the continuum of carechallenges hospitals to design and implement new medication management processes. With medication errors contributing to patient morbidity and mortality,1 establishing a comprehensive process for reconciling a patient's medications during the hospitalization episode is an important quality improvement and patient safety goal.

However, the current state of inpatient medication management is highly fragmented. Standard documentation is lacking, as is integration of information between care settings.2 There are now reports describing implementation of various medication reconciliation processes for admissions,3 transfers,4 and discharges.5

Hospitalists are well‐positioned to contribute to the implementation of medication reconciliation. Indeed, because TJC does not explicitly specify what type of health care provider (eg, physician, nurse, etc.) should assume responsibility for this process, institutions have designed workflows to suit their own needs, while striving to comply with national standards.

Given the complexity and lack of standardization around this NPSG, a survey was distributed to attendees of a Society of Hospital Medicine (SHM) national meeting to determine the various processes implemented thus far, and to ascertain existing challenges to implementation. We report here on the results.

METHODS

A survey tool (Appendix) was designed to query demographic and institutional factors, involvement in the process, and barriers to implementation of medication reconciliation. Surveys were included in all attendees' registration materials, resulting in the distributions of approximately 800 surveys.

Responses were entered into an Excel spreadsheet. Simple descriptive statistics were used to determine proportions for providers, processes, and barriers to implementation. Where appropriate, variables were dichotomized, allowing for paired t‐test analysis. Statistical significance was defined as a P value less than .05. Subgroup analyses by hospital type, provider type, and process method were performed.

RESULTS

A total of 295 completed surveys were collected. The responses are tabulated in Table 1.

DISCUSSION

Managing medication information for inpatients is an extremely complex task. On admission, home medication lists are often inaccurate or absent,6 requiring extra time and effort to discover this information. By discharge, medication regimens have frequently been altered,7 making communication of changes to the next provider essential. One study described myriad provider, patient, and health system issues in maintaining accurate outpatient medication lists.8 These issues are further compounded by the multiple prescribers, necessary hand‐offs, and formulary differences in the inpatient setting.

Over half of the hospitalists in this survey reported hospitalist involvement in design and implementation of medication reconciliation. Given the familiarity with hospital systems and inpatient workflow, hospitalists are well‐positioned to contribute to successful implementation. Nonetheless, many were unaware of efforts to implement this NPSG.

Measurement of both process and outcome measures is important when determining value in quality improvement. Beyond process measures, outcome measures such as adverse drug events, readmission rates, mortality, patient satisfaction, and outpatient provider satisfaction may be appropriate in evaluating medication reconciliation strategies. Even measuring the accuracy of the process with respect to the admission orders written would be a valuable source of information for further improvement. Unfortunately, respondents indicated that evaluation was occurring infrequently. Potentially more problematic is the apparent lack of clarity regarding identification of healthcare provider responsibility for specific process steps. By far the least uniformity is in the acquisition and documentation of the preadmission medication list. There is variability in who is assigned to perform this task, but a substantial number of respondents indicated that their process involved a shared responsibility between physicians and nurses. It is unclear whether this phenomenon reflects the complexity of inpatient medication information management, or is simply an attempt to distribute the work among providers. Sharing the work between physicians and nurses may increase the overall likelihood for compliance and possibly improve the safety and accuracy of the process, especially if the physicians and nurses take the medication history in a redundant fashion and share their findings. Conversely, compliance may decrease if each provider merely expects the other to complete the process. Optimally, an interdisciplinary workflow for medication history taking would be in place, involving both physicians and nurses, with the availability of pharmacist consultation in complex cases. However, our survey data suggest this is infrequent; resident physicians appear to be the ones shouldering substantial responsibility for medication reconciliation in tertiary academic centers. Further research into the accuracy of medication reconciliation processes involving different strategies for medication information collection would be useful.

We documented several barriers to successful implementation of medication reconciliation. Physicians cited a lack of medication knowledge on the part of the patient and unavailable prior medication lists as substantial barriers to success. Many medication reconciliation processes are limited by issues of poor health literacy or inadequate patient knowledge about medications. This lack of medication knowledge is especially problematic for patients new to a healthcare system. It will be important to implement processes that not only reconcile medications accurately, but also make medication information available for future care episodes.

Time required to complete the process was also important. Certain elements of the medication reconciliation process are new work, and integrating the process into existing workflows is crucial. Given the significant time commitment required, the rare involvement of pharmacists at most institutions is striking. It appears that hospital pharmacists do not currently own any of the medication reconciliation process steps at most facilities, despite having formal training in medication history‐taking. In the 2006 ASHP national hospital pharmacy survey, one‐third of pharmacists stated that there were not enough pharmacy resources to meet medication reconciliation demands; only 19% of those surveyed stated pharmacists provided medication education at discharge to more than 25% of their patients.9

This report has several limitations. The survey used was not comprehensive, and only represents a convenience sample of hospitalists attending anational meeting. Nearly 300 physicians responded, representing both teaching and private hospital settings. We consider the response rate of 37% reasonable for a survey of this nature, and the variety of processes described is likely indicative of the overall status of medication reconciliation implementation. The over‐representation of certain institutions in our survey is possible, especially those with large or influential hospital medicine programs. Our survey did not ask respondents to name their home institutions. In addition, this design is open to a convenience sample bias, in that surveying only national meeting attendees (rather than the entire SHM membership) risks overinclusion of those hospitalists involved in leadership roles and quality improvement projects. Despite this, the variety of processes described is likely indicative of the overall status of medication reconciliation implementation in mid‐2006. It is possible that processes have become more uniform nationwide in the interim.

Our survey results reflect the complexity surrounding medication reconciliation. It appears that full implementation has not yet occurred everywhere, significant barriers remain, and outcome measurement is limited. Importantly, physicians, nurses, and pharmacists do not have standardized roles. Responsibility for medication reconciliation has predominantly been added to the existing duties of inpatient physicians and nurses, with limited involvement of pharmacists. Hospitalists are well‐positioned to lead the ongoing implementation of medication reconciliation processes and should take advantage of their systems knowledge to effectively partner with other physicians, nurses, and pharmacists to achieve success in medication reconciliation.

The Joint Commission's (TJC) National Patient Safety Goal (NPSG) #8Accurately and completely reconcile medications across the continuum of carechallenges hospitals to design and implement new medication management processes. With medication errors contributing to patient morbidity and mortality,1 establishing a comprehensive process for reconciling a patient's medications during the hospitalization episode is an important quality improvement and patient safety goal.

However, the current state of inpatient medication management is highly fragmented. Standard documentation is lacking, as is integration of information between care settings.2 There are now reports describing implementation of various medication reconciliation processes for admissions,3 transfers,4 and discharges.5

Hospitalists are well‐positioned to contribute to the implementation of medication reconciliation. Indeed, because TJC does not explicitly specify what type of health care provider (eg, physician, nurse, etc.) should assume responsibility for this process, institutions have designed workflows to suit their own needs, while striving to comply with national standards.

Given the complexity and lack of standardization around this NPSG, a survey was distributed to attendees of a Society of Hospital Medicine (SHM) national meeting to determine the various processes implemented thus far, and to ascertain existing challenges to implementation. We report here on the results.

METHODS

A survey tool (Appendix) was designed to query demographic and institutional factors, involvement in the process, and barriers to implementation of medication reconciliation. Surveys were included in all attendees' registration materials, resulting in the distributions of approximately 800 surveys.

Responses were entered into an Excel spreadsheet. Simple descriptive statistics were used to determine proportions for providers, processes, and barriers to implementation. Where appropriate, variables were dichotomized, allowing for paired t‐test analysis. Statistical significance was defined as a P value less than .05. Subgroup analyses by hospital type, provider type, and process method were performed.

RESULTS

A total of 295 completed surveys were collected. The responses are tabulated in Table 1.

DISCUSSION

Managing medication information for inpatients is an extremely complex task. On admission, home medication lists are often inaccurate or absent,6 requiring extra time and effort to discover this information. By discharge, medication regimens have frequently been altered,7 making communication of changes to the next provider essential. One study described myriad provider, patient, and health system issues in maintaining accurate outpatient medication lists.8 These issues are further compounded by the multiple prescribers, necessary hand‐offs, and formulary differences in the inpatient setting.

Over half of the hospitalists in this survey reported hospitalist involvement in design and implementation of medication reconciliation. Given the familiarity with hospital systems and inpatient workflow, hospitalists are well‐positioned to contribute to successful implementation. Nonetheless, many were unaware of efforts to implement this NPSG.

Measurement of both process and outcome measures is important when determining value in quality improvement. Beyond process measures, outcome measures such as adverse drug events, readmission rates, mortality, patient satisfaction, and outpatient provider satisfaction may be appropriate in evaluating medication reconciliation strategies. Even measuring the accuracy of the process with respect to the admission orders written would be a valuable source of information for further improvement. Unfortunately, respondents indicated that evaluation was occurring infrequently. Potentially more problematic is the apparent lack of clarity regarding identification of healthcare provider responsibility for specific process steps. By far the least uniformity is in the acquisition and documentation of the preadmission medication list. There is variability in who is assigned to perform this task, but a substantial number of respondents indicated that their process involved a shared responsibility between physicians and nurses. It is unclear whether this phenomenon reflects the complexity of inpatient medication information management, or is simply an attempt to distribute the work among providers. Sharing the work between physicians and nurses may increase the overall likelihood for compliance and possibly improve the safety and accuracy of the process, especially if the physicians and nurses take the medication history in a redundant fashion and share their findings. Conversely, compliance may decrease if each provider merely expects the other to complete the process. Optimally, an interdisciplinary workflow for medication history taking would be in place, involving both physicians and nurses, with the availability of pharmacist consultation in complex cases. However, our survey data suggest this is infrequent; resident physicians appear to be the ones shouldering substantial responsibility for medication reconciliation in tertiary academic centers. Further research into the accuracy of medication reconciliation processes involving different strategies for medication information collection would be useful.

We documented several barriers to successful implementation of medication reconciliation. Physicians cited a lack of medication knowledge on the part of the patient and unavailable prior medication lists as substantial barriers to success. Many medication reconciliation processes are limited by issues of poor health literacy or inadequate patient knowledge about medications. This lack of medication knowledge is especially problematic for patients new to a healthcare system. It will be important to implement processes that not only reconcile medications accurately, but also make medication information available for future care episodes.

Time required to complete the process was also important. Certain elements of the medication reconciliation process are new work, and integrating the process into existing workflows is crucial. Given the significant time commitment required, the rare involvement of pharmacists at most institutions is striking. It appears that hospital pharmacists do not currently own any of the medication reconciliation process steps at most facilities, despite having formal training in medication history‐taking. In the 2006 ASHP national hospital pharmacy survey, one‐third of pharmacists stated that there were not enough pharmacy resources to meet medication reconciliation demands; only 19% of those surveyed stated pharmacists provided medication education at discharge to more than 25% of their patients.9

This report has several limitations. The survey used was not comprehensive, and only represents a convenience sample of hospitalists attending anational meeting. Nearly 300 physicians responded, representing both teaching and private hospital settings. We consider the response rate of 37% reasonable for a survey of this nature, and the variety of processes described is likely indicative of the overall status of medication reconciliation implementation. The over‐representation of certain institutions in our survey is possible, especially those with large or influential hospital medicine programs. Our survey did not ask respondents to name their home institutions. In addition, this design is open to a convenience sample bias, in that surveying only national meeting attendees (rather than the entire SHM membership) risks overinclusion of those hospitalists involved in leadership roles and quality improvement projects. Despite this, the variety of processes described is likely indicative of the overall status of medication reconciliation implementation in mid‐2006. It is possible that processes have become more uniform nationwide in the interim.

Our survey results reflect the complexity surrounding medication reconciliation. It appears that full implementation has not yet occurred everywhere, significant barriers remain, and outcome measurement is limited. Importantly, physicians, nurses, and pharmacists do not have standardized roles. Responsibility for medication reconciliation has predominantly been added to the existing duties of inpatient physicians and nurses, with limited involvement of pharmacists. Hospitalists are well‐positioned to lead the ongoing implementation of medication reconciliation processes and should take advantage of their systems knowledge to effectively partner with other physicians, nurses, and pharmacists to achieve success in medication reconciliation.

The Joint Commission's (TJC) National Patient Safety Goal (NPSG) #8Accurately and completely reconcile medications across the continuum of carechallenges hospitals to design and implement new medication management processes. With medication errors contributing to patient morbidity and mortality,1 establishing a comprehensive process for reconciling a patient's medications during the hospitalization episode is an important quality improvement and patient safety goal.

However, the current state of inpatient medication management is highly fragmented. Standard documentation is lacking, as is integration of information between care settings.2 There are now reports describing implementation of various medication reconciliation processes for admissions,3 transfers,4 and discharges.5

Hospitalists are well‐positioned to contribute to the implementation of medication reconciliation. Indeed, because TJC does not explicitly specify what type of health care provider (eg, physician, nurse, etc.) should assume responsibility for this process, institutions have designed workflows to suit their own needs, while striving to comply with national standards.

Given the complexity and lack of standardization around this NPSG, a survey was distributed to attendees of a Society of Hospital Medicine (SHM) national meeting to determine the various processes implemented thus far, and to ascertain existing challenges to implementation. We report here on the results.

METHODS

A survey tool (Appendix) was designed to query demographic and institutional factors, involvement in the process, and barriers to implementation of medication reconciliation. Surveys were included in all attendees' registration materials, resulting in the distributions of approximately 800 surveys.

Responses were entered into an Excel spreadsheet. Simple descriptive statistics were used to determine proportions for providers, processes, and barriers to implementation. Where appropriate, variables were dichotomized, allowing for paired t‐test analysis. Statistical significance was defined as a P value less than .05. Subgroup analyses by hospital type, provider type, and process method were performed.

RESULTS

A total of 295 completed surveys were collected. The responses are tabulated in Table 1.

DISCUSSION

Managing medication information for inpatients is an extremely complex task. On admission, home medication lists are often inaccurate or absent,6 requiring extra time and effort to discover this information. By discharge, medication regimens have frequently been altered,7 making communication of changes to the next provider essential. One study described myriad provider, patient, and health system issues in maintaining accurate outpatient medication lists.8 These issues are further compounded by the multiple prescribers, necessary hand‐offs, and formulary differences in the inpatient setting.

Over half of the hospitalists in this survey reported hospitalist involvement in design and implementation of medication reconciliation. Given the familiarity with hospital systems and inpatient workflow, hospitalists are well‐positioned to contribute to successful implementation. Nonetheless, many were unaware of efforts to implement this NPSG.

Measurement of both process and outcome measures is important when determining value in quality improvement. Beyond process measures, outcome measures such as adverse drug events, readmission rates, mortality, patient satisfaction, and outpatient provider satisfaction may be appropriate in evaluating medication reconciliation strategies. Even measuring the accuracy of the process with respect to the admission orders written would be a valuable source of information for further improvement. Unfortunately, respondents indicated that evaluation was occurring infrequently. Potentially more problematic is the apparent lack of clarity regarding identification of healthcare provider responsibility for specific process steps. By far the least uniformity is in the acquisition and documentation of the preadmission medication list. There is variability in who is assigned to perform this task, but a substantial number of respondents indicated that their process involved a shared responsibility between physicians and nurses. It is unclear whether this phenomenon reflects the complexity of inpatient medication information management, or is simply an attempt to distribute the work among providers. Sharing the work between physicians and nurses may increase the overall likelihood for compliance and possibly improve the safety and accuracy of the process, especially if the physicians and nurses take the medication history in a redundant fashion and share their findings. Conversely, compliance may decrease if each provider merely expects the other to complete the process. Optimally, an interdisciplinary workflow for medication history taking would be in place, involving both physicians and nurses, with the availability of pharmacist consultation in complex cases. However, our survey data suggest this is infrequent; resident physicians appear to be the ones shouldering substantial responsibility for medication reconciliation in tertiary academic centers. Further research into the accuracy of medication reconciliation processes involving different strategies for medication information collection would be useful.

We documented several barriers to successful implementation of medication reconciliation. Physicians cited a lack of medication knowledge on the part of the patient and unavailable prior medication lists as substantial barriers to success. Many medication reconciliation processes are limited by issues of poor health literacy or inadequate patient knowledge about medications. This lack of medication knowledge is especially problematic for patients new to a healthcare system. It will be important to implement processes that not only reconcile medications accurately, but also make medication information available for future care episodes.

Time required to complete the process was also important. Certain elements of the medication reconciliation process are new work, and integrating the process into existing workflows is crucial. Given the significant time commitment required, the rare involvement of pharmacists at most institutions is striking. It appears that hospital pharmacists do not currently own any of the medication reconciliation process steps at most facilities, despite having formal training in medication history‐taking. In the 2006 ASHP national hospital pharmacy survey, one‐third of pharmacists stated that there were not enough pharmacy resources to meet medication reconciliation demands; only 19% of those surveyed stated pharmacists provided medication education at discharge to more than 25% of their patients.9

This report has several limitations. The survey used was not comprehensive, and only represents a convenience sample of hospitalists attending anational meeting. Nearly 300 physicians responded, representing both teaching and private hospital settings. We consider the response rate of 37% reasonable for a survey of this nature, and the variety of processes described is likely indicative of the overall status of medication reconciliation implementation. The over‐representation of certain institutions in our survey is possible, especially those with large or influential hospital medicine programs. Our survey did not ask respondents to name their home institutions. In addition, this design is open to a convenience sample bias, in that surveying only national meeting attendees (rather than the entire SHM membership) risks overinclusion of those hospitalists involved in leadership roles and quality improvement projects. Despite this, the variety of processes described is likely indicative of the overall status of medication reconciliation implementation in mid‐2006. It is possible that processes have become more uniform nationwide in the interim.

Our survey results reflect the complexity surrounding medication reconciliation. It appears that full implementation has not yet occurred everywhere, significant barriers remain, and outcome measurement is limited. Importantly, physicians, nurses, and pharmacists do not have standardized roles. Responsibility for medication reconciliation has predominantly been added to the existing duties of inpatient physicians and nurses, with limited involvement of pharmacists. Hospitalists are well‐positioned to lead the ongoing implementation of medication reconciliation processes and should take advantage of their systems knowledge to effectively partner with other physicians, nurses, and pharmacists to achieve success in medication reconciliation.

Ventilator‐associated pneumonia (VAP) is a serious and common complication for patients in the intensive care unit (ICU).1 VAP is defined as a pulmonary infection occurring after hospital admission in a mechanically‐ventilated patient with a tracheostomy or endotracheal tube.2, 3 With an attributable mortality that may exceed 20% and an estimated cost of $5000‐$20,000 per episode,49 the management of VAP is an important issue for both patient safety and cost of care.

The diagnosis of VAP is a controversial topic in critical care, primarily because of the difficulty distinguishing between airway colonization, upper respiratory tract infection (eg, tracheobronchitis), and early‐onset pneumonia. Some clinicians insist that an invasive sampling technique (eg, bronchoalveolar lavage) with quantitative cultures is essential for determining the presence of VAP.10 However, other clinicians suggest that a noninvasive approach using qualitative cultures (eg, tracheal suctioning) is an acceptable alternative.11 Regardless, nearly all experts agree that a specimen for microbiologic culture should be obtained prior to initiating antibiotics. Subsequent therapy should then be adjusted according to culture results.

Studies from both Europe and North America have demonstrated considerable variation in the diagnostic approaches used for patients with suspected VAP.12, 13 This variation is likely a result of several factors including controversy about the best diagnostic approach, variation in clinician knowledge and experience, and variation in ICU management protocols. Such practice variability is common for many ICU behaviors.1416 Quality‐of‐care proponents view this variation as an important opportunity for improvement.17

During a recent national collaborative aimed at reducing health careassociated infections in the ICU, we discovered many participants were uncertain about how to diagnose and manage VAP, and considerable practice variability existed among participating hospitals. This uncertainty provided an important opportunity for developing consensus on VAP management. On the basis of diagnostic criteria outlined by the Centers for Disease Control and Prevention (CDC), we developed algorithms as tools for diagnosing VAP in 4 ICU populations: infant, pediatric, immunocompromised, and adult ICU patients. We also developed an algorithm for initial VAP treatment. An interdisciplinary team of experts reviewed the current literature and developed these evidence‐based consensus guidelines. Our intent is that the algorithms provide guidance to clinicians looking for a standardized approach to the diagnosis and management of this complicated clinical situation.

METHODS

Our primary goal was to develop practical algorithms that assist ICU clinicians in the diagnosis and management of VAP during daily practice. To improve the quality and credibility of these algorithms, the development process used a stepwise approach that included assembling an interdisciplinary team of experts, appraising the published evidence, and formulating the algorithms through a consensus process.18

RESULTS

DISCUSSION

In this article, we introduce algorithms for diagnosing and managing VAP in infant, pediatric, immunocompromised, and adult ICU patients. We developed 4 algorithms because the hospitals in our system care for a wide range of patients. Our definitions for VAP were based on criteria outlined by the CDC because these rigorously developed criteria have been widely disseminated as components of the Institute for Healthcare Improvement's ventilator bundle.120 Clinicians should be able to easily incorporate these practical algorithms into their current practice.

The algorithms were developed during a collaborative across a large national health care system. We undertook this task because many clinicians were uncertain how to integrate the enormous volume of VAP literature into their daily practice, and we suspected there was large variation in practice in our ICUs. Recent studies from other health care systems provided empiric evidence to support this notion.12, 13

We offer these algorithms as practical tools to assist ICU clinicians and not as proscriptive mandates. We realize that the algorithms may need modification based on a hospital's unique bacteriology and patient populations. We also anticipate that the algorithms will adapt to future changes in VAP epidemiology, preventive strategies, emerging pathogens, and new antibiotics.

Numerous resources are available to learn more about VAP management. An excellent guideline from the Infectious Diseases Society of America and the American Thoracic Society discusses VAP issues in detail,5 although this guideline only focuses on immunocompetent adult patients. The journal Respiratory Care organized an international conference with numerous VAP experts in 2005 and subsequently devoted an entire issue to this topic.81 The Canadian Critical Care Trials Group and the Canadian Critical Care Society conducted systematic reviews and developed separate guidelines for the prevention, diagnosis, and treatment of VAP.80, 121

In summary, we present diagnostic and treatment algorithms for VAP. Our intent is that these algorithms may provide evidence‐based practical guidance to clinicians seeking a standardized approach to diagnosing and managing this challenging problem.